SUBDUCTION TO STRIKE-SLIP TRANSITIONS ON PLATE BOUNDARIESpeople.uncw.edu/grindlayn/revabstr_vol.pdf · Penrose Conference: Subduction to Strike-Slip Transitions on Plate Boundaries,

Penrose Conference

SUBDUCTION TO STRIKE-SLIP TRANSITIONS ONPLATE BOUNDARIES

Puerto Plata, Dominican RepublicJanuary 18-24, 1999

ABSTRACT VOLUME

Conveners: Paul Mann, Nancy Grindlay and James Dolan

Sponsored by the Geological Society of America, Petroleum ResearchFund, NSF and NASA

Penrose Conference: Subduction to Strike-Slip Transitions on Plate Boundaries, Jan. 18-24, 1999

1

STRUCTURAL GEOLOGY AND SEDIMENTOLOGY OF A PLIOCENEINNER-TRENCH SLOPE SUCCESSION, NORTHWESTERN ECUADOR

K. R. AALTO

Dept. of Geology, Humboldt State Univ., Arcata, CA, USA 95521, [email protected]

The Pliocene Upper Onzole Formation exposed in the vicinity of Punta Gorda, nearEsmeraldas, Ecuador, is composed mainly of fine-grained mud turbidites, having regularvertical sequences of sedimentary structures associated with a positive grading, andbioturbation restricted mostly to the tops of beds. The remainder of beds measured consist ofvolcanic ash, mud pelagite, and glauconitic silt-sand turbidites. Vertical sequential analysis ofstratigraphic sections for the most part show no pronounced trends in bed thickness or grainsize. Volcanic ashes are crystal-vitric tuffs occurring in four bedding styles: A) normally-graded ashes with burrowed gradational tops and sharp wavy bases; B) ashes that form part ofa complex microstratigraphy consisting of thinly-bedded mudstone, silt-sand turbidites,tuffaceous mudstone, and ash beds; C) less conspicuous ash laminae and ash-filled burrows;and D) a tuffaceous bed with ash and mud swirled together in convolute layers. Ash chemistrysuggests an Andean high-K calc-alkaline provenance. Facies relations, paleontologic data andregional geologic setting suggest sediment accumulation on an inner trench slope in a basinsituated oceanward of the Pliocene trench-slope break.

Faulting is extensive and reflects two deformational episodes, the youngest involvingHolocene marine and fluvial terraces containing Pre-Columbian artifacts. Faults group into anolder, northerly-trending, listric set, and a younger, WNW-trending high-angle set. Faultstriations of both sets suggest dominantly dip-slip motion. Associated with the older set offaults is a spaced fracture cleavage defined by dark, curvilinear to sinuous S-surfaces that areless than 1 mm wide but tens-of-cm long, oriented subparallel to listric faults. Cleavagesurfaces are spaced over distances as great as several meters from larger faults. Individualsurfaces are defined in thin section by cataclastic deformation resulting in grain diminution,commonly characterized by small amounts of offset. The faults reflect an episode of Quaternarytrench-normal extension in the forearc, followed by trench-parallel extension. Both sets offaults crosscut large NE-trending regional folds.

Daly (1989) documented east-west extensional deformation among late Oligocene-middleMiocene rocks in northwest Ecuador, but did not refer to similar strain recorded amongPliocene rocks. This deformation resulted in uplift of the southern and eastern margins of theBorbón Basin in the middle Miocene, which resulted in influx of coarse clastics into the basinfrom the east. Evans and Whittaker (1982) recognize successive episodes of uplift to the souththat resulted in the influx of coarse turbidite clastics into the basin during the early Pliocene.Following deposition of these turbidites, mudstone deposition and basin deepening resumed,and the portion of the Upper Onzole exposed at Punta Gorda accumulated. It is possible thatthe episode of extensional deformation that produced the older, north-trending faults in thestudy area reflects reactivation of faults in the accretionary wedge that were responsible for theearlier episodes of Neogene basin inversion proposed by Evans and Whittaker (1982). Daly(1989) relates trench-normal extension in the forearc to decreasing plate-convergence rates,which resulted in collapse of the forearc wedge. Although he does not specifically relate aNeogene rate decrease to an episode of extensional deformation, this decrease may help toexplain the genesis of the older, north-trending faults observed in the study.

During the Neogene the landward margin of the Ecuador trench may have also suffered nettectonic erosion, which is an alternative hypothesis to explain Neogene east-west extensionrecorded by the first generation of faults in the study area. If Neogene basin evolution in theEcuadorian forearc reflects trench-normal extension, the resulting basins may be structurally


2

and morphologically similar to those of the sediment-rich Central American forearc, a regionalso characterized by block-faulting reflecting extension orthogonal to the Middle AmericaTrench. Another analog for the depositional setting of the Upper Onzole is the outer deep-marine basin portion of the terraced compound forearc basin of the Alaskan-Aleutian Arc,although these outer basins formed in response to transpressional and transtensionaldeformation associated with strike-slip faulting in the forearc (Ryan and Scholl, 1989). Noconcrete evidence of strike-slip faulting exists in either the study area. Daly (1989) invokesoblique convergence-induced strike-slip-related deformation in the Ecuadorian forearc duringthe late Miocene to Recent to explain a variety of regional structures. However, the stressvectors associated with his deformational fields (resulting in NE-SW crustal shortening andNW-SE extension) are incompatible with what is recorded in the Esmeraldas region, unless thestructures reflect a complex partitioning of strain within a discrete block.

The younger WNW-trending high-angle normal faults reflect NNE-SSW extension, withfaults oriented generally orthogonal to the modern Ecuadorian trench, a deformational fieldreflected in the orientation of some of the large faults on the regional map. Both these faultsand the north-trending set of faults shown on the map truncate the large regional folds.Following Upper Onzole deposition, changes in stress regime, from generally NW-SEcompression, to E-W tension, to NNE-SSW tension are difficult to explain. Perhaps thisprogressive deformation reflects stress rotation that resulted from the encroachment andcollision of the leading edge of the continent with the thick crust of the Carnegie Ridge (Daly,1989). This change in stress fields would have produced a final episode of “basin inversion”in the Esmeraldas region.

Daly, M. C., 1989. Correlations between Nazca/Farallon plate kinematics and forearc basin evolution in Ecuador. Tectonics 8 , 769-790.

Evans, C. D. R., and Whittaker, J. E., 1982. The geology of the western Part of the Borbón Basin, northwest Ecuador. In: Trench-Forearc Geology (edited by J. K. Leggett). Geological Society of London Special Publication 10 , 191-198.

Ryan, H. F., and Scholl, D. W., 1989. The evolution of forearc structures along an oblique convergent margin, central Aleutian arc. Tectonics 8 , 497-516.


Overview of structures associated with subduction to strike-slip transitions in NewZealand

P.M. Barnes1, J-C. Audru2,5, J-Y. Collot3, B. Davy4, J. Delteil2, R. Herzer4, G. Lamarche3,1,J-F. Lebrun3, B. Mercier de Lépinay2, R. Sutherland4, R. Wood4.

1National Institute of Water and Atmospheric Research (NIWA), PO Box 14901, Kilbirnie, Wellington, NewZealand.2UMR Géosciences Azur, UNSA-CNRS, Valbonne, 06050, Sophia Antipolis, France.3UMR Géosciences Azur, ORSTOM, 06235, Villefranche sur Mer, France.4Institute of Geological and Nuclear Sciences, PO Box 30368, New Zealand.5Presently at Institut de Physique du Globe, 67084 Strasbourg cedex, France

The structure of the Pacific-Australian plate boundary in the NewZealand region varies along its lengthin response to changes in crustalstructure, interplate coupling, andobliquity of relative motion betweenthe plates. Two opposite-facingsubduction to strike-slip transitionsoccur on the plate boundary, one ateach end of the South Island of NewZealand, where oblique subduction ofoceanic crust in the north and southmerges with an oblique continentalcollision zone in central South Island(Fig. 1). The northern transition occursfrom the west-dipping Hikurangisubduction zone, through thecontinental strike-slip MarlboroughFault System, to the oblique dextralAlpine Fault collision zone. Thesouthern transition occurs from thealpine collision zone to the east-dipping Puysegur – Fiordlandsubduction system south of NewZealand.

Fig. 1. Two subduction to strike-slip transitions inthe New Zealand region.

Both regions of tectonic transition have existed as subduction to strike-slip transition plateboundaries since the Miocene. However, the geometry of the boundaries, the activity ofgeological structures, and the kinematics of faulting have changed through time as relativeplate motion changed and as the northern part of the plate boundary rotated clockwise. Thetransition regions are presently undergoing oblique convergence at a relative plate motion rateof 3.5-4.0 cm per year.


During the last five years several complementary research teams have utilised deep-crustaland shallow-penetration seismic reflection data, shipborne gravity and magnetic data,EM12Dual multibeam swath imagery, MR1 side-scan sonar, 3.5 kHz profiles, seabed samples,and onland field data to reveal the active structures of both the northern [1,2,3] and southern[4,5,6] subduction to strike-slip transition areas. The Hikurangi – Marlborough transition straddles an oceanic - continental boundary in thesubducting Pacific Plate crust, and a consequent southward increase in interplate couplingacross the southern end of the subduction zone. Thickened oceanic crust (12-15 km) is beingsubducted beneath North Island, where geological strain is being partitioned incompletelybetween an axial belt of strike-slip faults and a largely contractional forearc. To the south,where the edge of continental Pacific Plate has been subducted beneath northeastern SouthIsland, almost all of the late Quaternary plate motion is recorded in the upper plate principallyacross a distributed zone of strike-slip faults that branch off the Alpine Fault and extendoffshore from Marlborough into the southern end of the Hikurangi margin. Significant changesin the shallow crustal structure of the transition region during the last 2 million years maypartially reflect the locking up of the southern end of the subduction zone. The Fiordland – Puysegur transition straddles a change from a continental plate boundary insouthwestern New Zealand to an intra-oceanic transform boundary on the Puysegur Ridge.The transition region evolved from a continental rift connected with an oceanic spreadingsystem in the Eocene and Oligocene, to its present day strike-slip – trench – strike-slipcharacter. The location of the Alpine Fault is well constrained on the Fiordland margin. Thesouthern end of the fault terminates abruptly at the trench deformation front and is alignedwith inherited structures on the rifted edge of a continental sliver on the Australian Plate. Onepossible model is that the abrupt transfer of plate motion from the Alpine Fault to thesubduction zone may be facilitated by a tear in the subducted plate developing along aninherited tectonic structure. Contractional strain along the Fiordland margin is beingpartitioned west of the fault onto lower slope accretionary wedges that are developing underextremely oblique convergence. At a first order scale the Puysegur Trench subductiondécollement together with a distributed zone of transpressive faults crossing the trench slopesouth of New Zealand transfer most of the plate motion across a mega-scale left step from theAlpine Fault to the oceanic strike-slip Puysegur Fault. The most important difference between the two strike-slip to subduction transitions in NewZealand is the distribution of upper plate strike-slip deformation through the transfer region. Inthe northern transition the transfer of plate motion is accommodated across numerous strike-slip and oblique-slip structures, whereas in the south the margin parallel motion is morelocalised. This results largely from differences in the crustal structure and coupling of theplates, and in the geological evolution of each region. An important commonality to bothtransitions is the late Neogene capturing of parts of the Pacific Plate in response to changes inplate motion kinematics.

Selected References:[1] Collot, J-Y. et al., 1996, From oblique subduction to intra-continental transpression:structures of the southern Kermadec-Hikurangi margin from multibeam bathymetry, side-scansonar and seismic reflection. Marine Geophysical Researches, 18: 357-381.[2] Audru, J-C, 1996, from the Hikurangi subduction to the Alpine Fault, Marlborough, NewZealand, Phd Thesis, University of Nice, France.[3] Barnes, P.M. et al., 1998, Strain partitioning in the transition area between oblique subduction and continental collision:Hikurangi margin, New Zealand. Tectonics, 17, 534-557.


[4] Collot, J-Y. et al., 1995, Morphostructure of an incipient subduction zone along atransform plate boundary: Puysegur ridge and trench, Geology, 23, 519-522.[5] Delteil, J. et al., 1996, From strike-slip faulting to oblique subduction: A survey of theAlpine Fault – Puysegur Trench transition, New Zealand, results of Cruise Geodynz-sud Leg2, Marine Geophysical Reseachers, 18, 383-399.[6] Lebrun, J-F., 1997, From the Alpine Fault to the Macquarie Ridge Complex, developmentof the puysegur subduction, southwest New Zealand, PhD Thesis, University of Paris Vi,France.


6

Upper plate Quaternary strain rates, kinematics and paleoseismology in thetransition from Hikurangi subduction to Southern Alps collision, New Zealand

Kelvin Berryman

New Zealand Institute of Geological & Nuclear Sciences, PO Box 30368, Lower Hutt. ([email protected])

To appreciate the present-day role of upper plate structures in the transition from the Hikurangisubduction margin to Southern Alps collision a synopsis of the late Neogene evolution of the regionis useful. Key factors or observations that bear upon the present-day snapshot are:• Total strike-slip motion on faults of the Marlborough system range from c. 2-34 km (Little &

Jones, 1998), excluding the Wairau fault which has c. 140 km of total dextral slip (Mazengarb etal., 1993) - most of it presumably when it acted as the single northern continuation of the Alpinefault prior to the development of the Marlborough fault system.

• Total late Neogene dextral slip on the Wellington and Wairarapa faults in the North Island appearsto be no more than a few kilometres (Begg & Mazengarb, 1996).

• The combination of total slip and current slip rate data imply that the Marlborough faults haveaccommodated dextral motion for much longer than the current dextral faults in southern NorthIsland.

• In the late Miocene there was a broad forearc basin in southern North Island perhaps similar tothat in present-day Sumatra (Cashman et al., 1992; Kelsey et al., 1995).

• Beginning in the early Pliocene (and perhaps earlier in the region of the present-day coastalranges) there was broad regional uplift. This may coincide with the westward march ofsubduction of the overthick Hikurangi plateau (Kelsey et al., 1995).

• A marine strait existed at the western margin of the forearc basin until the early Pleistocene.• The frontal ridge comprising the Tararua and Ruahine ranges did not become prominent features

of the North Island until the last c. 1 Myr.• The change from pure strike-slip motion on the Alpine fault to a component of convergence and

resultant growth and erosion of the Southern Alps c. 5-7 Myr ago (Walcott, 1998) has had adramatic impact on the volume of the sedimentary prism offshore of Marlborough and southernNorth Island. Changes in the internal dynamics of the subduction margin can therefore beexpected, and may have resulted in a change in the location where the partitioned translationalcomponent of motion will be accommodated.

While the upper plate kinematics in the northern South Island may have been reasonably constant inthe past few million years (apart from possible diminishing slip on the Wairau fault, and inceptionand development of the Porters Pass-Amberley fault system as the newest Marlborough fault), therehave been rapid changes in the location, style, and rate of deformation in tectonic domains ofsouthern North Island. About 80±20% of the translational component of plate motion is presentlyaccommodated on 3-4 dextral strike-slip faults in southern North Island (Beanland, 1995). Thesehave average dextral slip rates that range from 1-10 mm/yr, have single event displacements thatrange from c. 4-15 m, and average recurrence intervals for these maximum magnitude surface faultrupture events of c. 350-5000 yrs (Van Dissen & Berryman, 1996). The associated earthquakemagnitudes are probably in the range Mw 7-8. Only the Wairarapa fault has ruptured in the 160 yearhistorical period (in 1855 AD) in a Mw 8 event. The zone of dextral faulting is located 100-160 kmfrom the trench. At distances of 60-100 km from the trench there is a relatively stable region (currently, but there isevidence of significant late Miocene-Pliocene contraction in this region)(Kelsey et al., 1995), andthen there are a system of coastal ranges which appear to be deforming rapidly by west-dippingreverse or oblique reverse faults immediately offshore of the coast. Coastal uplift of 1-4 m as aconsequence of rupture of the coastal zone of faults has recurrence intervals of 500-2000 years


7

(Berryman et al., 1989; Berryman, 1994). In Marlborough the orientation of the principal faults of the Marlborough fault belt are within a fewdegrees of the relative plate motion vector, and the sum of the 10-20 kyr averaged slip rates is withinerror of the NUVEL 1A plate motion rate (Holt & Haines, 1995). The majority of the slip is on thesouthern-most, well-developed fault - the Hope fault. Slip rates on this one fault are c. 20 mm/yr,about one-half of the relative plate motion (Van Dissen & Yeats, 1991; Cowan & McGlone, 1991). Suggestions of an episodic clustering of upper plate large earthquakes in the coastal area of theHikurangi margin (Berryman et al., 1989) appears to extend to northeastern South Island where aminor component of upper plate shortening is taken up by shortening on structures between theprincipal strike-slip faults (Ota et al., 1996; Miyauchi et al., in prep). Thus, upper plate clustering oflarge earthquake activity appears to extend across the transition from an active subduction zone toMarlborough where all of the relative plate motion is in the upper plate and the former subductionzone is regarded as fossil (Reyners & Cowan, 1993).

Beanland, S. 1995. The North Island dextral fault belt. PhD thesis, Victoria University of Wellington.Begg, J.G.; Mazengarb, C. 1996. Geology of the Wellington area, scale 1:50,000. Institute of Geological & Nuclear

Sciences geological map 22. 1 sheet+128 p. Lower Hutt: Institute of Geological & Nuclear Sciences Ltd.Berryman, K.R. 1994. Age, height and deformation of Holocene marine terraces at Mahia Peninsula, Hikurangi

subduction margin, New Zealand. Tectonics 12: 1347-1364.Berryman, K.R.; Ota, Y.; Hull, A.G.; 1989: Holocene paleoseismicity in the fold and thrust belt of the Hikurangi

subduction zone, eastern North Island, New Zealand. Tectonophysics 163: 185-195.Cashman, S.M.; Kelsey, H.M.; Erdman, C.F.; Cutten, H.N.C.; Berryman, K.R. 1992: A Structural Transect and

Analysis of Strain Partitioning across the Forearc of the Hikurangi Subduction Zone, Southern Hawke's Bay,North Island, New Zealand. Tectonics 11: 242-257.

Cowan, H.A.; McGlone, M.S. 1991: Late holocene displacements and characteristic earthquakes on the Hope Riversegment of the Hope fault, New Zealand. Journal of the Royal Society of New Zealand 21(4): 373-384.

Holt, W.E.; Haines, A.J. 1995. The kinematics of northern South Island, New Zealand, determined from geologicstrain rates. Journal of Geophysical Research 100 (B9): 17991-18010.

Kelsey, H.M.; Cashman, S.M.; Beanland, S.; Berryman, K.R. 1995. Structural evolution at the inboard margin of theaccretionary wedge, Hikurangi forearc, New Zealand. Tectonics 14: 1-18.

Little, T.A.; Jones, A. 1998. Seven million years of strike-slip and related off-fault deformation, northeasternMarlborough fault system, South Island, New Zealand. Tectonics 17: 285-302.

Mazengarb, C.; Begg, J.G.; Simes, J.E. 1993. Does the Esk Head subterrane cross Cook Strait (abs) Misc. Pub. Geol.Soc. N.Z. 79A: 113.

Miyauchi, T.; Brown, L.; Ota, Y.; Fujimori, T.; Berryman, K. in prep. Age, height and tectonic significance of lateHolocene marine terraces along south Marlborough coast, New Zealand.

Ota, Y.; Pillans, B.; Berryman, K.R.; Beu, A.; Fujimori, T.; Miyauchi, T.; Berger, G. 1996. Pleistocene coastalterraces of Kaikoura Peninsula and the Marlborough coast, South Island, New Zealand. New Zealand Journal ofGeology and Geophysics 39: 51-73.

Reyners, M., and H. Cowan, The transition from subduction to continental collision: crustal structure in the NorthCanterbury region, New Zealand, Geophys. J. Int., 115, 1124-1136, 1993.

Van Dissen, R. J., and R. S. Yeats, Hope fault, Jordan thrust, and uplift of the seaward Kaikoura range, New Zealand,Geology, 19, 393-396, 1991.

Walcott, R.I. 1998. Modes of oblique compression: Late Cenozoic tectonics of the South Island of NewZealand. Reviews of Geophysics 36: 1-26.


8

Seismicity of Two Adjacent but Geometrically Different Subduction to Strike-Slip Transition Zones: San Cristobal and North New Hebrides Trenches,southwest Pacific

Andria Bilich1,2, Cliff Frohlich2, Paul Mann2

1 Department of Geological Sciences; University of Texas at Austin; Austin, Texas 78712-11012 Institute for Geophysics; 4412 Spicewood Springs Road #600; University of Texas at Austin; Austin, Texas78759-8500

OverviewThis study outlines the similarities and differences between two subduction to strike-

slip transition zones possessing different plate intersection geometries (closed corner versusopen corner trench orientation). We apply seismicity analyses to two transition zones of thesouthwest Pacific: the oblique subduction regime of the southern Solomon Islands and theclassic subduction regime of Vanuatu. These two zones are geographically adjacent, as the SanCristobal Trench of the Solomon Islands becomes the North New Hebrides Trench near SantaCruz (Vanuatu), and offer interesting points of contrast.

We employ four different data sources in this study, three earthquake seismicitycatalogs and a compilation of local geologic data. The Harvard centroid moment tensor (CMT)catalog includes focal mechanisms over a 21-year period, giving short-term tectonic andmoment information. The Engdahl catalog of relocated hypocenters covers 31 years and wasused for information dependent on earthquake location, such as seismic frequency trends anddepths. We also used the Abe historical catalog for M > 7.0 events between 1897 and 1976 toenhance moment calculations for these transition zones. The compilation of geologic data is aseries of maps synthesizing onshore geology, faults, folds, and bathymetry. Seismic trends: frequency, depth, and slip vectors

Two trends of hypocenter depth and frequency clearly delineate the two separatetransition zones. Intermediate-depth seismicity (70 – 300 km) is present only in the subductionportions of the San Cristobal and North New Hebrides Trenches, and all events in thetransform section connecting the two subduction zones is shallower than 50 km. Earthquakefrequency follows similar trends, with 3 to 4 times more earthquakes occurring in thesubduction zones than in the transform zone. We observe this frequency trend in both theHarvard CMT and Engdahl catalogs.

Drawing boundaries between the subduction, transform, and trench-orientationchangeover areas, we created 4 geographic regions used for comparative analysis.Interestingly, these geographic regions closely correspond to abrupt changes in slip vectordomains, i.e. primary slip vectors are perpendicular to the trench in subduction zones, paralleland curve with the trench at the closed corner of Santa Cruz, and are parallel to subparallel inthe open corner (Solomon Islands) and transform regions.

With both slip vector domains and average seismicity trends, the secondary slip vector(the slip vector described above is the primary slip vector) is controlled by the plate motionvector, as demonstrated by roughly parallel strikes of the second slip vector and the localNUVEL-1 plate motion. Therefore, both trench orientation (dictating breaking stresses) andplate motion direction play important roles in determining earthquake type at different sectionsof the trench. Seismic trends: mechanisms, moment, and consistency

We constructed ternary diagrams [see Cliff Frohlich’s abstract for explanation] andmoment-weighted average focal mechanisms to get a handle on the overall seismic trends in thefour regions. The ternary diagrams clearly show a scarcity of strike-slip mechanismsthroughout the study area, with strike-slip mechanisms conspicuously absent from the


9

transform region. Average focal mechanisms (a moment-weighted summation of all HarvardCMTs for a region) show strong thrust character in the Solomon Islands and Vanuatu regions,but only a small strike-slip influence is added in for the transform and closed corner transitions.The ternary diagrams support this trend, as well as show the mixed nature of transform regionseismicity (thrust, normal, and strike-slip earthquake types in relatively equal amounts).

Trends in seismically-expressed moment roughly parallel the consistency trendsoutlined above. The earthquake moment per kilometer per year in the subduction zones is 5 to8 times greater than that of the transform region. The transform region’s shallow, infrequentseismicity does not energetically expresses the local tectonics. Geologic data

We also compiled fault and fold orientation data from Hackman, Petterson, and theBritish Geological Society in order to relate the geology to stress orientations and seismicactivity. An immediate clue to stress orientations are the two sets of perpendicular faults andfolds predominantly on San Cristobal. Fold axis orientations trend WNW throughout theisland, where these fold axes are paried with two main fault trends—one almost perpendicularand trending NNE, and one parallel set of fault orientations. Many of the synthetic faultsstretch across the width of the island, while antithetic faults extend over much shorterdistances.

Tying these observations to trench orientation, we draw a correlation between a highrate of seismic activity and a lack of faults, folds, convoluted topography, or bathymetricexpression. Also, the main fault and fold trends of Guadalcanal and San Cristobal are muchmore consistent with local trench orientations, not with the fairly constant subduction direction.


10

AMS dating of sediment samples from the Dominican Republic

G.S. Burr1, Carol Prentice2, Paul Mann3, Martitia Tuttle4, andJ. McGeehin5

1NSF-Arizona AMS Facility, University of Arizona, Physics Department, Tucson, AZ 857212USGS, 345 Middlefield Road, MS 977, Menlo Park, CA 940253Institute for Geophysics, University of Texas at Austin, 4412 Spicewood Springs Rd. Bldg. 600, Austin, TX 787124University of Maryland, Department of Geology, College Park, MD 207425Radiocarbon Laboratory, USGS National Center, Reston, VA 22092

Radiocarbon dating is widely used in paleoseisomological studies to calculate the age ofindividual strata and determine rates of motion along faults. Such rate estimates depend on theidentification of stratigraphic layers, their relative displacement in three dimensions, and onaccurate age determinations for each sediment layer. In the optimal case one could collectcoarse (>few mm) pieces of charcoal or wood from every stratigraphic layer of interest andanalyze them for 14C. In practice such cases do not normally occur. Alternative candidates for 14C dating include bulk sediments, terrestrial gastropods, pollen,and soil organic matter (SOM). All of these are inferior to wood or charcoal dates. It is well-known for example, that organics in soils move at different rates, which vary from annual tomillennial timescales (Jenkinson and Raynor, 1977; Dörr and Münnich, 1986; Trumbore andZheng, 1996). The problem with dating SOM is to identify an organic fraction in the soilwhich was introduced at the time the sediment was deposited and has not been chemicallyaltered since deposition. To address this problem we have measured the radiocarbon content ofa variety of different SOM fractions using both physical and chemical organic carbon separates.Methods employed include sieving, differential thermal combustion and combinations of acidand base pretreatments. The suite of dates from the SOM fractions allows us to assess thepotential variability and uncertainties in the 14C ages of the soils. A second potential problemfor SOM dates stems from the likely occurrence of physical contamination in the tropicalenvironment (Scharpenseel and Becker-Heidmann, 1992). Ultimately the usefulness of SOMdates depends on a thorough understanding of these processes, which are specific to each siteand soil type. Our eventual goal is to use the radiocarbon results from the SOM fractions inconjunction with those measured from charcoal or wood samples collected from the samehorizons to document the usefulness of the SOM dates on their own. In addition to the bulk sediment and SOM results we have made measurements on severalspecies of terrestrial gastropods. It is shown that these samples have a reservoir correctionwhich is related to the species of gastropod. We have also observed a gradient in the amountof 14C from the exterior to the interior of the gastropod shells, suggesting post-depositionalexchange of carbon. We use a selective dissolution pretreatment developed for corals toremove this secondary contamination (Burr et al., 1992).

References

Burr, G.S., R.L. Edwards, D.J. Donahue, E.R.M. Druffel and F.W. Taylor. 1992. Massspectrometric 14C and U-Th measurements in coral. Radiocarbon 34(3): 611-618.

Dörr, H. and K.O. Münnich. 1986. Annual variations of the 14C content of soil CO2. In:Stuiver, M. and R.S. Kra, eds. Proceedings of the 12th International 14C Conference.Radiocarbon 31(3): 264-268.


11

Jenkinson, D.J. and J.H. Raynor. 1977. The turnover of SOM in some of the Rothamstedclassical experiments. Soil Science 123: 298-305.

Scharpenseel, H.W. and P. Becker-Heidmann. 1992. Twenty-five years of radiocarbondating soils: paradigm of erring and learning. Radiocarbon 34(3):541-549.

Trumbore, S.E. and S. Zheng. 1996. Comparison of fractionation methods for soil organicmatter 14C analysis. Radiocarbon 38(2): 219-229.


Plate boundary zone deformation in the northern Caribbean: Observations,measurements and models

Eric Calais

CNRS Géosciences Azur – 250 Rue A. Einstein – 06560 Valbonne – France ([email protected])

The Northern Caribbean plate boundary (NCPB) mostly consists of a system of left-lateral strike-slip faultsconnected to the west with the middle America subdction zone, to the east with the Lesser Antilles subductionzone. In its western part, the NCPB trace is the purely strike-slip Swan fault, marine extension of the Polochic-Motagua fault zone of Central America (Rosencrantz and Mann, 1991). The Swan fault connects to the eastwith the Oriente through the Mid-Cayman spreading center, a short oceanic spreading ridge that has beengenerating the oceanic crust of the Cayman trough since late Eocene (Rosencrantz et al., 1988). From the Mid-Cayman spreading center to the western edge of the Cuban margin, the geometry of the Oriente fault is poorlyknow but earthquake focal mechanisms indicate pure strike-slip motion. Along the western part of the Cubanmargin, the Oriente fault is a system of sinistrally offset en échelon fault segments associated with pull-apartbasins. In contrast, the central and eastern parts of the Cuban margin are characterized by activecompressional structures (échelon folds, reverse faults and southverging thrusts) that form a 300 km longnarrow zone of transpressive deformation along the Oriente fault (Calais and Mercier de Lépinay, 1991). It isassociated with frequent moderate earthquakes, with focal mechanisms showing either pure thrust or acombination of thrust and strike-slip (Perrot et al., 1997; Calais et al., 1998). To the east, in the WindwardPassage, the Oriente fault is associated with active en échelon folds and flower structures (Calais and Mercierde Lépinay, 1995). Further east, the Oriente fault follows the northern Haitian coastline and connects with theSeptentrional fault in northern Hispaniola (Mann et al., 1984).

The Septentrional fault is responsible for the uplift of the Cordillera Septentrional in the northernDominican Republic and for active folding and faulting at its contact with the Cibao valley (Edgar, 1989; DeZoeten and Mann, 1991; Calais et al., 1992; Prentice et al., 1993). Offsets in terraces dated by radiocarbon ongastropods shells suggest that the Holocene slip rate across the fault is less than 13 ± 4 mm/yr (Mann et al.,1998). Marine geophysical surveys north of Hispaniola (Dillon et al., 1992; Dolan et al., 1998) and PuertoRico (Masson and Scanlon, 1991; Speed and Larue, 1991; Grindlay et al., 1997) indicate that theSeptentrional fault zone extends to the east north of the island of Puerto Rico with seismic evidences for strike-slip motion in that area (McCann and Sykes, 1984; Calais et al., 1992). These surveys have also revealedactive compressional tectonic structures along the northern Hispaniola margin (folds and north-vergingthrusts), that might be associated with some of the largest earthquakes along that segment of the northernCarribean plate boundary (Russo and Villasenor, 1995; Dolan and Wald, 1998). These compressive structuresconnect to the east with the Puerto Rico trench, where the subduction of the Atlantic oceanic lithosphere isassociated with a south-dipping Benioff zone and a very strong negative gravity anomaly (-400 mGals). Northof the Puerto Rico trench, the available structural data show a significant subsidence of its inner wall alonglarge normal faults (Le Pichon et al., 1985; Heezen et al., 1985; Masson and Scanlon, 1991; Speed and Larue,1991). Finally, active faulting associated with a significant seismicity has also been identified along theEnriquillo fault zone in southern Hispaniola and along the Muertos acretionnary prism south of the DominicanRepublic and Puerto Rico. At the longitude of the Dominican Republic anc Puerto Rico, the relative motion ofthe Caribbean plate with respect to the North American plate is therefore accomodated by a 200 km widedeformation zone undergoing left-lateral shear along the Septentrional and Enriquillo faults, limited to thenorth and to the south by two large thrusts with opposite vergence in the Muertos and Northern Hispaniola-Puerto Rico trenches.

This wealth of structural and seismotectonic data has prompted the development of a number of models toexplain the present-day deformation along the NCPB. Among the most recent ones, Deng and Sykes (1995)emphasize the role of compressive structure and thrusts earthquakes and assume a highly convergent motionbetween the Caribbean and North American plates. Calais and Mercier de Lépinay (1993) and Russo and

Penrose Conference: Subduction to Srike-Slip Transitions on Plate Boundaries, Jan. 18-24, 1999

Villasenor (1995) emphasize the role of strike-slip structures and earthquakes, and assume a mostly strike-slipmotion between the Caribbean and North American plates. In such models, compressive structures are relatedto the geometry of the major strike-slip faults. Mann et al. (1995) and Lundgren and Russo (1996) emphasizethe role of independant microplates within the plate boundary zone. Dolan and Wald and Mann et al. (1998)emphasize the role of strain partitioning of the mainly east-west interplate motion into onshore/nearshorestrike-slip faults (Septentrional and Enriquillo faults) and thrust faults (Muertos trough and northernHispaniola margin). In the absence of quantitative data and direct measurements of the strain distributionwithin the plate boundary zone, these models are difficult to test. Indeed, the strain distribution among theactive structures of the NCPB is not yet quantified and the slip rate along the major active faults is not known.In addition, the relative motion of the Caribbean and north American plates is poorly defined in global platemotion models because of a lack of constraining data such as transform azimuths, ocean ridge spreading ratesaround the Caribbean plate. Moreover, slip vectors of earthquakes in its northeastern part are likely to bebiased by strain partitioning in a kinematic setting where strike-slip and convergence coexist. GlobalPositioning System (GPS) measurements in the northeastern Caribbean (CANAPE project) are now starting toshed some light on these issues.

In 1986, a sparse network of geodetic sites was occupied by GPS in the northeastern Caribbean, as part of aNASA program to test and validate this new technology in a humid tropical environment (Dixon et al., 1991).In 1994, this network was reoccupied and a number of new sites were added, in particular in the DominicanRepublic. Some of these sites have been reoccupied in 1995, 1996, and 1998. 27 new sites have beenestablished in the Dominican Republic in September 1998, they will be occupied during a two-week campaignat the end of January 1999. The analysis of the GPS data collected in 86, 94, and 95 shows an eastward motionof the Caribbean plate at a rate of 17 mm/yr relative to North America, measured at Cabo Rojo, southernDominican Republic (Dixon et al., 1998). Additional measurements performed at 4 sites in the DominicanRepublic in September 1998 confirm this result, with slightly faster velocities (18 to 21 mm/yr in a N80Eazimuth). Similar velocities have been found on Aves island, confirming a velocity of the Caribbean platerelative to North America about twice higher that the NUVEL-1A prediction of 11(3 mm/yr. We modeled thevelocity field measured at six sites in the Dominican Republic and one on Grand Turk Island on the BahamaPlatform using three-dimensional dislocations in an elastic half-space. The fault geometry and seismogenicdepth were taken for geological and geophysical observations and imposed. Although this type of forwardmodelling by trial and error is non-unique, the tests that we have performed so far require a far-field NSconvergent velocity field superimposed on the mostly eastward displacement of the Caribbean plate. Thisconvergence is mainly accomodated on the northern Hispaniola fault zone, which ruptured in a series ofmagnitude 7-8 earthquakes in the 1940s and accomodates oblique slip at a rate of 5 mm/yr (4 mm/yr strike-slip and 3 mm/yr dip-slip for the best fit model). The models indicate that the Muertos thrust accomodates 3 to1 mm/yr of pure NS convergence. They show a slip rate of 9 mm/yr on the Septentrional fault (8 mm/yr strike-slip and 2 mm/yr dip-slip for the best fit model), in agreement with the geologicaly-derived Holocene rate,and 13 mm/yr of pure strike-slip on the Enriquillo fault. Trenching across the Septentrional fault indicates thatit last ruptured 730 years ago, slipping about 5 m horizontally and 2 m vertically (Prentice et al., 1993). At aslip rate of 9 mm/yr, this fault has accumulated about 6 m of potential slip since that last rupture, which couldcause a M>7.5 earthquake if it was enterely released today (Wells and Coppersmith, 1994). In addition, wefind that the extraction of rigid rotations from the GPS-derived velocity field does not support the idea of acounter-clockwise rotation of the Puerto Rico block as proposed in some models but indicate east-westextension between Hispaniola and Puerto Rico.

These recent but still preliminary GPS results, together with the most recent tectonic and seismotectonicdata, seem therefore to favor a model with a Caribbean/North America plate motion slightly oblique to theplate boundary trace and partitioned into strike-slip faulting in Hispaniola and along the inner wall of thePuerto Rico trench and thrust faulting offshore north and south of Hispaniola and Puerto Rico.

References :Calais E., N. Bethoux, and B. Mercier de Lépinay, From transcurrent faulting to frontal subduction: A

seismotectonic study of the northern Caribbean plate boundary from Cuba to Puerto Rico, Tectonics, 11, 114-123, 1992.

Penrose Conference: Subduction to Strike-Slip Tansitions on Plate Boundaries, Jan. 18-24, 1999

Calais, E., and B. Mercier de Lépinay, From transtension to transpression along the northern Caribbean plateboundary off Cuba: Implications for the recent motion of the Caribbean plate, Tectonophysics, 186, 329-350,1991.

Calais, E., and Mercier de Lépinay, B. Semi-quantitative modeling of strain kinematics relations alongtranscurrent plate boundaries: application to the present-day motion of the Caribbean plate relative to NorthAmerica. J. Geophys. Res., 98, 8293-8308, 1993.

Calais, E., and Mercier De Lépinay, B. Strike-slip processes in the Northern Caribbean between Cuba andHispaniola (Windward Passage). Marine Geophysical Researches, 17, 63-95, 1995.

Calais, E., Perrot, J., and Mercier de Lépinay, B., Strike-slip tectonics and seismicity along the NorthernCaribbean plate boundary from Cuba to Hispaniola, Geol. Soc. of Amer. Special Paper, 1998.

Deng, J., and L.R. Sykes, Determination of Euler pole for contemporary relative motion of Caribbean andNorth American plates using slip vectors oof interplate earthquakes, J. Geophys. Res., 85, 2669-2678, 1995.

De Zoeten, R., and P. Mann, Structural geology and Cenozoic tectonic history of the central CordilleraSeptentrional, northern Dominican Republic, in Geologic and Tectonic Development of the North America-Caribbean Plate Boundary Zone in Hispaniola, edited by P. Mann, G. Draper and J. Lewis, Spec. Pap. Geol.Soc. Amer., 262, 265-280, 1991.

Dillon, W.P., Austin, J.A., Scanlon, K., M., Edgar, N.T., and Parson, L.M., Accretionary margin of north-western Hispaniola: Morphology, structure, and development of the northern Caribbean plate boundary,Marine and Petroleum Geology, 9, 70–92, 1992.

Dixon, T., Farina, F., DeMets, C., Jansma, P. ,Mann, P., and Calais, E., 1998, Relative motion between theCaribbean and North American plates and related boundary zone deformation based on a decade of GPSobservations, J. Geophys. Res., 103, 15157-15182, 1998.

Dixon, T.H., G. Gonzalez, S.M. Lichten and E. Katzigris, First epoch geodetic measurements with the GlobalPositionning System across the Northern Caribbean plate boundary zone, J. Geophys. Res., 96, 2397-2415,1991.

Dolan, J.F., and Wald, D.J., Strain accumulation and seismic energy release localized along collisionalasperities: the 1946 and 1948 north-central Caribbean earthquakes, Geol. Soc. Amer. Spec. Paper, 1998.

Edgar, N.T., Structure and geological development of the Cibao Valley, northern Hispaniola, paper presentedat 12th Caribbean Geology Conference, Ste Croix, Virgin Islands, 1989.

Heezen, B.C., W.D. Nesteroff, M. Rawson, and R.P. Freeman-Lynde, Visual evidence for subduction in thewestern Puerto Rico trench, in Symposium Géodynamique Caraïbe, 287-304, Technips, Paris, 1985.

Ladd, J.W., J.L. Worzel, and J.S. Watkins, Multifold seismic reflection records from the northern Venezuelabasin and north slope of Muertos trench, in Island Arcs, Deep Sea Trenches and Back-Arc Basins, MauriceEwing Ser., vol.1, edited by M. Talwani and W.C. Pitman III, pp. 41-56, AGU, Washington D.C., 1977.

Le Pichon, X., J. Iiyama, J. Bourgois, B. Mercier de Lépinay, J. Tournon, C. Muller, J. Butterlin, and G.Glaçcon, Premiers résultats de la campagne d'essais du submersible françcais Nautile dans la fosse de PortoRico (Grandes Antilles), C.R. Acad. Sci., 301, 743-749, 1985.

Lundgren, P.R., and R.M. Russo, Finite element modeling of crustal deformation in the north America-Caribbean plate boundary zone, J. Geophys. Res., 1996

Mann P., Burke K. and Matumoto T., Neotectonics of Hispaniola plate motion, sedimentation and seismicityat a restraining bend. Earth and Planetary Science Letters, 70, 311-324, 1984.

Mann, P., E. Calais, C. DeMets, T. Dixon, J. Dolan, N. Grindlay, P. Jansma, and C. Prentice, Integratedgeologic and geophysical data suggest high seismic risk for the Dominican Republic, Haiti, and Puerto Rico,EOS, submitted, 1998.

Mann, P., F.W. Taylor, R. Edwards, and T.L. Ku, Actively evolving microplate formation by oblique collisionand sideways motion along strike-slip faults : An example from the Northeastern Caribbean plate margin,Tectonophysics, 246, 1-69, 1995.

Masson, D.G., and K.M. Scanlon, The neotectonic setting of Puerto Rico, Geol. Soc. Am. Bull., 103, 144-154,1991.

McCann, W.R., and L.R. Sykes, Subduction of aseismic ridges beneath the Caribbean plate: Implications forthe tectonic and seismic potential of the northeastern Caribbean, J. Geophys. Res., 89, 4493-4519, 1984.

Perrot, J., Calais, E., and Mercier de Lépinay, B., Waveform simulation of the May 25th, 1992, Ms=6.7 CaboCruz earthquake (Cuba): new constraints on the tectonic regime along the northern Caribbean plateboundary. PAGEOPH, 149, 475-487, 1997.


Prentice, C., P. Mann, F.W. Taylor, G. Burr, and S. Valastro, Paleoseismicity of the North American plateboundary (Septentrional fault), Dominican Republic, Geology, 21, 49-52, 1993.

Rosencrantz, E., and P. Mann, SeaMarc II mapping of transform faults in the Cayman Trough, Caribbean Sea,Geology, 19, 690-693, 1991.

Rosencrantz, E., M.I. Ross, and J.G. Sclater, The age and spreading history of the Cayman Trough asdetermined from depth, heat flow and magnetic anomalies, J. Geophys. Res., 93, 2141-2157, 1988.

Russo, R.M., and Villasenor, A., The 1946 Hispaniola earthquakes and the tectonics of the North America-Caribbean plate boundary zone, northeastern Hispaniola, J. Geophys. Res., 100, 6265-6280, 1995.

Speed, R.C., and D.K. Larue, Extension and transtension in the plate boundary zone of the northeasternCaribbean, Geophys. Res. Lett., 18, 573-576, 1991.

Sykes, L.R., McCann, W.R., and Kafka, A.L., Motion of the Caribbean plate during the last 7 million yearsand implications for earlier Cenozoic movements, J. Geophys. Res., 87, 10,656-10,676, 1982.

Wells, D.L., and K.J. Coppersmith, New empirical relationships among magnitude, rupture length, rupturewidth, rupture area, and surface displacement, Bull. Seism Soc. Am., 84, 974-1002, 1994.


16

ANALISIS DE FOTOLINEAMIENTOS DE LA REPUBLICA DOMINICANA

Sergio Chiesa (*), Giovanna Civelli (°), Simone De Toni (+).

(*) CNR-Centro di Studio per la Geodinamica Alpina e Quaternaria, Milano, Italia. E-mail [email protected](°) Geotec arl - Via Artigiani, 1, Brusaporto (Bg), Italia. E-mail [email protected](+) Geoisa International S.A. - San Francisco Dos Rios, San Josè, Costa Rica. E-mail [email protected].

In this poster we present the results of the study of one of the argument in the project"Proyecto D: Prevención de Riesgos Geológicos (Riesgos Sísmicos) del Programa deDesarrollo Geológico-Minero SYSMIN", financed by the European Union.

At the workplace, we performed the followings activities:• An analysis of somes LANDSAT TM images (Thematic Mapper) at scale 1:250.000,

covering all the national territory of the Dominican Republic.• Some in sito surveys aimed to look for evidences of neotectonics.• The drafting of one computerized map at scale 1:500.000 of all the national territory which

shows all the lineaments identified by LANDSAT images.• A statistical study of the radial distribution and the lineament persistence.

The term "lineament" corresponds to all the linear elements, rectilinear or curvilinear, whicheffects the topographic surface and which can be related to the geological and particularytectonic phenomena. The study of LANDSAT images allows to draw attention to thedeformation of rigid style, which is tipically caracterized by high angles, while deformations ofplastic style, often with low angles, may not result too clear.

We have subdivided the Dominican Republic into 12 homogeneous zones, according to the"terranes" recognized in Hispañiola by Mann et al. (1991). The overall analysis has covered17200 lineaments, with a total lenght of 26000 km.

Some linear elements have been prevalently identified, with an average lenght of 2-3 km andmaximum lenght of 4-5 km. The spatial distribution of lineaments is naturally heterogeneous,with different density according to the various"terranes".

The mayor lineaments' density is shown on very important tectonics elements (fault zone)which in fact are formed by a field of minor elements with some fractures that follow the lineardevelopment of the fault zone; the width of this fault zone is related to mechanical caracteristicsof crossed rocks.

The analysis of lineaments' distribution shows the presence of a high concentration along abelt with a N-S trend between San Josè Ocoa, Bonao and Jima Abajo. This belt cuts off,breaks off or deviates a great number of lineaments with orientation E-W, NW-SE and NNW-SSE, which are those prevalent in the whole Dominican Republic and in each "terrane"; the beltmay represents a segment of the Gonave microplate margin.

Therefore, this structural element has an evident sense of neotectonic and might be the reasonof sismicity in the Bonao-Azua area.


17

Dominican Republic (Hispañiola Island, north-eastern Caribbean): a map ofmorpho-structural units at a scale 1:500000 through LANDSAT TM imageprocessing

Sergio Chiesa (*), Guido Mazzoleni (°)

(*) CNR–Centro di Studio per la Geodinamica Alpina e Quaternaria, Milano, Italia. Email [email protected](°) Geoisa Internacional S.A., San Francisco Dos Rios, S. José, Costa Rica. Email [email protected]

The territory of the Dominican Republic is located in an active plate margin area, at theboundary between the North American and the Caribbean plates. This area has been affected bycatastrophic earthquakes, both recently and in the past. For this reason, and with the objectiveof designing an instrumental network for monitoring the microseismicity of the area, all thegeological data available in the literature was collected and synthesised, and an original remotesensing study was carried out. The latter supplied results that agree well with the generalmorpho-structural framework described by different authors and confirmed some of theirdetails; in addition, it highlighted some innovative elements. The morpho-structural units thatwere recognised by this study seem to be in good agreement with the morpho-tectonic zonesproposed by Lewis (1980) and Lewis & Draper (1990).

The Northern Cordillera and the Samanà Peninsula are sharply interrupted to the south(towards Cibao Valley) by an evident boundary, which is probably both the most importanttectonic element and a seismic source, the Septentrional Fault Zone. This boundary is formedby a set of closely spaced morphological steps, which become progressively more evident fromwest to east.

The Eastern Peninsula appears to be a separated physiographic element, with the littledeformation present being mostly of the shear type, as no clear evidence of thrusting wasobserved. This area seems to be affected by a slight, progressive and homogeneous tectonicuplift. In the literature the peninsula is considered an area with minimal or null uplift ifcompared with the areas to the west of a structural line striking N-S and located 70°W to thewest (Mann et al., 1995). The remote sensing study presented in this paper detected a highlydeformed belt, striking roughly N-S, which extends from Boca de Yàsica to the north to theRìo Ocoa lower valley to the south. This is a slightly more western location than the structuralelement mentioned above.

By contrast, the Central Cordillera shows the morphological characteristics of an area that isaffected by fast and significant recent tectonic uplift. This is evidenced by the widespread slopeinstability phenomena on the steep mountain-sides and by the deeply cut drainage network.This morpho-structural unit appears to be clearly separated from the neighbouring tectonicvalleys, which, on the contrary, are strongly subsiding. The Central Cordillera and the EasternPeninsula join together through the most complex and composite sector in the whole areastudied. In this sector most of the transitional units identified are present. Some of these unitsshow clear indications of extensive/transtensive tectonics, evidenced by NW-SE orientedmorphological steps and horst/graben structures.

The eastern portion of the important lowland belt that contains Enriquillo Valley, the Lagunadel Rincòn, appears to be more affected by extension/transtension tectonics. The shape andgeometry of the morphological steps observed could indicate that this area is situated at thenorthern boundary of a pull-apart basin.


18

The Southern Penìnsula shows characteristics that partly recall the Eastern Península.However, it is clearly more deformed by both transtensive and thrusting tectonics, the latter inthe northern sector. Ridges, valleys and general morphological patterns, mostly orientedN50°W were observed. They are parallel to or form a low angle with extension/transtensionstructures, averagely striking N40°W, such as morphological steps, normal faults and propergrabens, like the most evident one near Pelempito.

References:Lewis J.F.(1980): Resumé of the geology of Hispañiola. In Field guide to the 9th Caribbean

Geological Conference, Santo Domingo, Dominican Republic, pp.5-31.Lewis J.F., Draper G.(1990): Geology and tectonic evolution of the northern Caribbean

margin. In Dengo G. & Case J.E. eds., The caribbean region. Boulder, Colorado,Geological Society of America, The Geology of North America, v. H, pp. 77-140.

Mann P., Taylor F.W., Edwards R.L., Ku-Teh-Lung (1995): Actively evolving microplateformation by oblique collision and sideways motion along strike-slip faults: en example fromthe northeastern Caribbean Plate margin. Tectonophysics 246, 1-3, pp. 1-69.


19

TERTIARY TECTONICS OF THE HISPANIOLA FAULT ZONE IN THENORTHWESTERN PIEDMONT OF THE CORDILLERA CENTRAL,DOMINICAN REPUBLIC

COLEMAN, Andrew Jay, ([email protected]), WINSLOW, Margaret Anne([email protected])

CITY COLLEGE OF NEW YORKDepartment of Earth and Atmospheric Sciences, J-106137th Street & Convent AvenueNew York, New York 10031

ABSTRACT New road-cuts and dam projects create a unique opportunity to produce a new geologic mapof the northwestern piedmont of the Cordillera Central (Figure 1). The Hispaniola Fault Zone(“HFZ”) crosses the piedmont region north of the Cordillera Central in the map area. Thisstudy area is bounded by the coordinates 70°45’W to 71°15’W and 19°30’N to 19°20’N withinthe San José de las Matas and Monción quadrangles. Field evidence confirms that strike-slipfaulting occurred along the HFZ during the middle to late Tertiary. Sinistral faults displaceOligocene conglomerates and Miocene sandy mudstones and limestones. Strike-slip faultmovement along the HFZ created a pull-apart basin as early as the Oligocene. The faultingpattern within the rocks of the pull-apart basin can be explained by a Riedel model for left-slipmovement. En-echelon folding and faulting within the pull-apart basin are consistent withregional deformation and the tectonic evolution of the Caribbean/North American plateboundary. New evidence from en-echelon faults on the north side of the HFZ suggests thatfault movement continued through the Late Miocene and quite possibly into the Early Pliocene. TECTONIC SETTING The study area along the northwestern segment of the HFZ is bounded by the Cibao Valley tothe north and by the Cordillera Central to the south (Figure 1). The HFZ displaces Tertiaryclastic rocks deposited along the dissected plateau (Palmer, 1979). This study focuses on thestructural geometry and evolution of the deformed Tertiary rocks. A Riedel model for left-slipmovement and new structural and stereonet data aid in this interpretation. The Tertiary rocks deposited in the pull-apart basin are folded and faulted in a structurallycomplex manner. Local structures in the field area include left-lateral strike slip faults,stretched pebbles and shear fractures. Tectonic breccias are also found along fault traces in thefield area. Data analyzed from these structures suggest that the pull-apart basin formed consistent with aRiedel model for left-slip movement. The Riedel model predicts that the pull apart basinevolved along the HFZ. Tectonic breccias found along the main faults of the HFZ evidencethat movement occurred from at least the Early Oligocene through the Early Miocene. Newtectonic breccias found along en echelon faults in the HFZ indicate that deformation occurredinto the Late Miocene. The structural development and movement along the faults of the pull-apart basin are related to the predominant trend of motion along the on-shore NorthernCaribbean Plate Boundary Zone given its current transpressional relative motion whichcombines transform slip as well as convergence (Molnar and Sykes, 1969; Sykes and others,1982, Pindell and Barrett, 1990; Mann and others, 1991).

Mann, P., Draper, G., and Lewis, J.F. 1991. An overview of the Geologic and TectonicDevelopment of Hispaniola: in Mann, P., Draper, G., and Lewis J.F. (eds), Geologic andTectonic Development of the North America-Caribbean Plate Boundary in Hispaniola.Geological Society of America Special Paper, 262, 1-28.


20

Molnar, P. and Sykes, L.R. 1969. Tectonics of the Caribbean and Middle America Regionsfrom Focal Mechanisms and Seismicity. Geological Society of America Bulletin, 80, 1639-1684. Palmer, H.C. 1979. Geology of the Monción-Jarabacoa Area, Dominican Republic in Lidz,B., and Nagle, F., (eds), Hispaniola; Tectonic focal point of the northern Caribbean; Threegeological studies in the Dominican Republic. Miami Geological Society, 29-68. Pindell, J.L. and Barrett, S.F. 1990. Geological evolution of the Caribbean region; A platetectonic perspective: in Dengo, G., and Case, J.E., (eds), The Geology of North America, TheCaribbean region. Geological Society of America, H, 405-432. Sykes, L.R., McCann, W.R., and Kafka, A.L. 1982. Motion of the Caribbean plate duringthe last 7 million years and implications for earlier Cenozoic movements. Journal ofGeophysical Research, 87, 10656-10676.

Figure 1 Location of the study area and general geologic features of the region.


21

Displacement partitioning and arc-parallel extension: Structural analysis of theAleutian arc transitional margin.

John Comstock and Hans G. Avé Lallemant

Rice University, Department of Geology and Geophysics, 6100 S. Main St., Houston, TX 77005

The Aleutian volcanic arc is an ocean to ocean, concave, convergent margin between theNorth American and Pacific plates along which convergence obliquity ranges fromperpendicular to the trench in the east to parallel to the trench in the west. Several authors(Jarrard, 1986; Geist et al., 1988; McCaffrey, 1991, 1992, 1996; Avé Lallemant, 1996) havedeveloped the hypothesis that this increase in convergence obliquity produces strain in the fore-arc that is partitioned into trench normal and trench parallel components that correspond to thetrench normal and parallel components of the convergence vector. Additionally, the margin ismost accurately described by a system of the two plates and a portion of the overriding platethat is a quasi-independent of both plates, a trans-contractional terrane. The transcontractionalterrane lies between the trench and back-arc bounding strike slip systems and is both translatedwest and internally extended as the convergence vector increases obliquity and rate. As thisvector approaches parallelism with the Aleutian trench, an increasing percentage of theincreasing convergence rate comprises the arc-parallel strike-slip component of thedisplacement. This differential displacement gradient necessarily produces internal extensionof the terrane (Avé Lallemant and Guth, 1990; McCaffrey, 1996).

Structural geology studies on Unalaska, Adak, and Attu Islands in the Aleutian arc confirmthat the deformational structures are not related directly to the relative plate convergence vector,but to trench normal and parallel components of the vector. All observed deformation wasbrittle, none ductile. Based on crosscutting and overprinting field relationships, data fromUnalaska (53N, 167W) fall into five groups: reverse faults (F1) related to NW-SE (arc-normal)contraction; conjugate strike-slip systems (F2) and (F3), related to the arc-parallel component ofconvergence and arc-parallel extension; normal faults (F4) perpendicular to the arc resultingfrom arc-parallel extension; and veins and dikes (F5), generally the youngest, showing arc-parallel stretching. Data from Adak (52N, 177W) show similar structures and arc-relativeorientations, although the absolute attitudes are rotated from the Unalaska data (commensuratewith the change in trench attitude.) Reverse and normal faults are less prominent, and thereforerelative timing is not as well constrained as on Unalaska. Additionally, Adak has NE-SWtrending, dextral splays between the arc-parallel dextral strike-slip systems, with individualfault blocks showing possible clockwise vertical-axis rotation. Data from Attu (53N, 173E)also show similar structures and arc-relative orientations, again with absolute attitudes rotatedclockwise from the data on Unalaska and Adak.

Analysis of the brittle structures on three islands across almost the entire length of theAleutian arc reveal a predominance of strike-slip and normal fault systems indicating arc-parallel extension and reverse fault structures indicating arc-normal contraction. Relative to thearc, the orientation of nearly all structures is constant. However, relative to the convergencevector of the North American and Pacific plates, the orientation of the structural elements ofthese Aleutian Islands varies considerably. The degree of penetrative deformation increases tothe west as well, implying greater strain as more of the plate convergence vector is transferredto strike-slip and the commensurate internal extension.

Unfortunately, due to poor exposure and lack of stratigraphic control, the magnitude ofdisplacement will be very difficult, if not impossible, to measure by conventional field geologicmethods. Part of our continuing research into the Aleutian arc displacement is a three yearGPS study, which should produce quantitative measurements after the 2000 field season.


22

Avé Lallemant, H.G., 1996, Displacement partitioning and arc-parallel extension in theAleutian volcanic island arc: Tectonophysics, v. 256, p. 279-293.

Avé Lallemant, H.G., and Guth, L.R., 1990, Role of extensional tectonics in exhumationof eclogites and blueschists in an oblique subduction setting, northwest Venezuela:Geology, v. 18, p. 950-953.

Geist, E.L., Childs, J.R., and Scholl, D.W., 1988, The origin of summit basins of theAleutian Ridge: Implications for block rotation of an arc massif: Tectonics, v. 7,p. 327-341.

Jarrard, R.D., 1986, Terrane motion by strike-slip faulting of forearc slivers: Geology, v.14, p. 780-783.

McCaffrey, R., 1991, Slip vectors and stretching of the Sumatran Forearc: Geology, v. 19,p. 881-884.

McCaffrey, R., 1992, Oblique plate convergence, slip vectors, and forearc deformation:Journal of Geophysical Research, v. 97, n. B6, p. 8905-8915.

McCaffrey, R., 1996, Estimates of modern arc-parallel strain rates in fore arcs: Geology,v. 24, p. 27-30.


23

Evidence for subduction locking and strain partitioning from GPSmeasurements on the southern Hikurangi Margin, New Zealand

Desmond Darby and John Beavan

Institute of Geological & Nuclear Sciences, PO Box 30-368, Lower Hutt, New Zealand ([email protected], [email protected])

The degree of locking of the Hikurangi subduction interface beneath the Wellington region ofNew Zealand is not well known. The Wellington and Wairarapa faults in the over-riding plateare the most important of many Quaternary transcurrent faults that are reactivated reverse faults.Interface locking has therefore occurred in the past, and would play a major part in determiningthe present interaction between the subduction interface and these crustal faults, and thepotential for great earthquakes in the region. A GPS project to address these problems started in 1992 with a small network crossing theWellington fault. This network was re-surveyed and extended in 1994 to cross the entiresouthern North Island. The larger network has since been re-surveyed in 1995 and 1996. Theaverage station spacing is about 12 km, and this is as great a density of GPS sites as has beenestablished anywhere. There are about 50 points in common to at least two of these foursurveys. Results show that the relative velocity across the 80-km network width is about one half theNUVEL-1A prediction for the Pacific and Australian Plates. Inversion of the velocities andtheir full variance-covariance matrix can determine some of the elastic dislocation parametersthat describe the source of the surface deformation as a locked segment on the subductioninterface. The locking occurs between depths of 6±1 km and 22±1 km. This depth rangecoincides with a paucity of seismicity, and its limits correspond with low-angle thrustearthquake focal mechanisms that have recently been obtained by Reyners et al. [1997]. Thefraction of plate motion presently locked is not very well determined, but probably lies between75-100%. The width over which a great thrust earthquake may occur is constrained to be about70 km down-dip. As velocities have become better determined through time, a concentration of strain hasbecome increasingly obvious across the Wellington fault trace. This is consistent withmeasurements obtained from a 5-m aperture rod strain-meter. Paleoseismic data indicate thisfault has a mean recurrence interval of 500-800 yr for M=7.5 events, with the last being 300-400 yr ago [Van Dissen and Berryman, 1996]. There is no corresponding concentration ofstrain across the sub-parallel Wairarapa fault to the east. This fault has a 1-2 kyr meanrecurrence interval for the last 4-5 events, and last ruptured in an M=8.2 earthquake in 1855,probably without slip on the plate interface [Darby and Beanland, 1992]. Forward modelling ofslipping and locked patches on both the subduction interface and the Wellington fault canprobably explain the GPS observations. However, algorithms to formally invert the data arebeing prepared, as our experience shows that inclusion of the covariance matrix is important. It remains to be seen whether fault listricity can be determined from the closely spacedgeodetic points that we have in this network. Although the deeper limit of locking lies verticallybeneath the Wellington fault, the geometry of these reactivated reverse faults may show thatthis limit can be better associated with the junction of the Wairarapa fault and the subductioninterface.


24

Multichannel seismic profiles showing deformation of the North Americanplate and plate boundary zone near the northeastern corner of the Caribbeanplate

William P. Dillon1, Elazar Uchupi2, Uri S. ten Brink1, Juan Acosta3, Christopher H. Anton1,Myung W. Lee4

1U.S. Geological Survey, Woods Hole MA, 02543, USA, [email protected], [email protected],[email protected] Hole Oceanographic Institution, Woods Hole, MA 02543, USA, [email protected] Español de Oceanografía, Corazon de Maria, #8 - 1°, 28002, Madrid, Spain, [email protected]. Geological Survey, Box 25046, Denver Federal Center, MS 960, Lakewood Colorado 80225, USA,[email protected]

The North American plate (NOAM) is moving westward relative to the Caribbean plate(CARIB), with which it is in contact. NOAM is being subducted along the eastern boundaryof CARIB, forming the Lesser Antilles island arc, and NOAM is moving past the northern sideof CARIB with a dominantly transcurrent motion. At the northeastern corner of CARIB, thepattern is complicated by the presence of a plate boundary zone (PBZ) that extends along thenorthern side of CARIB, between CARIB and NOAM. The PBZ consists of one or moremicroplates that support eastern Hispaniola, Puerto Rico and the Virgin Islands. The contactbetween PBZ and NOAM occurs at the Puerto Rico Trench, north of the islands, and thecontact between PBZ and CARIB occurs at the Muertos Trough, south of the islands. WhereNOAM is overridden by the CARIB and PBZ, it is forced into a compound bend so that it dipswestward to the east of the PBZ/CARIB and southward to the north of those plates. Thuscomplex stresses are being applied to contort the plate to its observed configuration.

Newly available multichannel seismic profiles are located on the deformed part of theNOAM and PBZ north of the Virgin Islands in an area of dominantly strike-slip motion. Thedata consist of a set of profiles collected jointly by the Instituto Español de Oceanografía andthe U.S. Geological Survey in 1997 and newly-released seismic data that were collected byShell International in 1974 (Dillon et al. 1998). Both sets of data were processed recently bythe USGS seismic processing center.

Multibeam bathymetry, which was also collected during the 1997 cruise, discloses ~100-km-long, en echelon, WNW-oriented steps on the NOAM near its contact with the PBZ.Seismic profiles show these as fault blocks, including horsts and grabens. Such extensionalfeatures were presumably formed by the stretching that resulted from bending of the NOAM asit was overridden by PBZ lithosphere; segmentation of the steps also may be partially the resultof the required extension caused by the compound warping of the North American plate. Theseismic profiles show the continuation of NOAM crust that has been subducted and that dipsbeneath the PBZ. The subducted NOAM lithosphere also shows fault-block steps, apparentlyparallel to those exposed north of the trench. It seems unlikely that these steps would bepreserved if significant subduction motion were normal to the plate boundary (normal to thePuerto Rico Trench axis), but their preservation is consistent with a dominantly transcurrentmotion along the plate boundary, in which the subduction occurs to the east and motionbetween plates is basically strike-slip.

In the central part of the PBZ, shallow, flat-lying Tertiary limestone strata, as young asPliocene age (Schneidermann et al., 1972; Moussa et al., 1987), form the Virgin Islandsplatform. However, on the north side of the PBZ, these strata have undergone major tilting (to4°) and substantial subsidence. This post-mid-Pliocene collapse has resulted in more than 5 kmof subsidence of platform rocks in the last ~3.3-3.6 my. Tilting of stratified limestone andactive earthquake shaking have resulted in detachment and sliding of limestone slabs along the


25

northern margin of Puerto Rico and the Virgin Islands (EEZ-SCAN 85 Scientific Staff, 1987;Schwab et al., 1991; Scanlon and Masson, 1996). Abrupt movement of these slabs may havegenerated some of the historical tsunamis near Puerto Rico and the Virgin Islands.Nondefinitive indications of small, offscraped tectonic accretionary wedges at the northern limitof the PBZ may suggest a minor component of shortening across the trench.

Two major tectonic movements have affected the region. The first is the westwardmovement of NOAM relative to CARIB (and the PBZ), which was initiated by a platereorganization in Eocene time and still continues. This is the dominant plate motion that is wellknown and that accounts for the subduction at the Lesser Antilles island arc and the basictranscurrent motion along the northern Caribbean plate boundary. The second major tectonicmovement that affects the northeastern Caribbean region is much younger, but it has resulted inmajor vertical movements. This is the post-mid-Pliocene set of movements that has resulted inmore than 5 km of subsidence in some locations. This motion probably accounts for thepresent depth of the Puerto Rico Trench system, which therefore, must be very young,consistent with the very small amount of sediment accumulated there. The extremely largefree-air gravity anomaly north of Puerto Rico (-380 mGal, the largest negative free-air gravityanomaly in the world) would logically be associated with the stresses that have resulted indownward crustal movements. The cause(s) of the catastrophic recent subsidence is not clear.However, analysis of seismicity patterns suggests that NOAM and CARIB, which are bothbent down under the PBZ, may be in contact at depth (Dillon et al. 1996). One could speculatethat some sort of restraining bend might exist at depth beneath the PBZ, such that thetranscurrent motion between NOAM and CARIB is converted to vertical motions.

REFERENCES

Dillon, W. P., Acosta, Juan, Uchupi, Elazar, ten Brink, U. S., 1998, Joint Spanish-American researchuncovers fracture pattern in northeastern Caribbean, EOS, Trans Am Geophysical Union, vol. 790, no. 28,p. 336-337.

Dillon W. P., Edgar, N. T., Scanlon, K. M., and Coleman, D. F., 1996, A review of the tectonic problems ofthe strike-slip northern boundary of the Caribbean Plate and examination by GLORIA, in, Gardner, J. V.,Field, M. E., and Twichell, D. C., eds., Geology of the United States’ Seafloor: The View from Gloria,Cambridge University Press, Cambridge, U.K., p. 135-164.

EEZ-SCAN 85 Scientific Staff, 1987, Atlas of the U.S. Exclusive Economic Zone, Eastern Caribbean. U.S.Geological Survey Miscellaneous Investigations Series I-1864 B, 58p., scale 1:500,000.

Moussa, M. T., Seiglie, G. A., Meyerhoff, A. A., and Taner, I., 1987, The Quebradillas Limestone ( Miocene-Pliocene), northern Puerto Rico, and tectonics of the northeastern Caribbean margin. Geological Society ofAmerica Bulletin, vol.99, p. 427-439.

Scanlon. K. M. and Masson, D. G., 1996, Sedimentary processes in a tectonically active region: Puerto RicoNorth Insular Slope, in, Gardner, J. V., Field, M. E., and Twichell, D. C., eds., Geology of the UnitedStates’ Seafloor: The View from Gloria, Cambridge University Press, Cambridge, U.K., p. 123-134.

Schneidermann, N., Beckmann, J. P., and Heezen, B. C., 1972, Shallow water carbonates from the Puerto RicoTrench Region. Transactions of the 6th Caribbean Geological Conference Volume, Margarita, Venezuela,p.423-425.

Schwab, W.C., Danforth, W.W., Scanlon, K.M., and Masson, D.G., 1991, A giant slope failure on thenorthern insular slope of Puerto Rico: Marine Geology, vol. 96, p. 237-246.


26

Strike-Slip to Subduction Transition in South-Central Alaska

Diane Doser, Monique Velasquez, Annette Velleiux and Rodolfo Lomas

University of Texas at El Paso

South-central Alaska is located in an area of complex deformation between the Pacific andNorth American plates. Northwestard motion of the Pacific plate relative to North Americaresults in subduction of the Pacific plate along the Aleutian trench west of 146 degrees W, andright-lateral strike-slip faulting along the Queen Charlotte/Fairweather fault system east of 136degrees W. The plate boundary is complicated by the collision of the Yakutat block, anallochthonous terrane, with North America. Study of the source parameters of magnitude > 5.5earthquakes occurring in this region since 1920 has been used to help determine how slip ispartitioned along this boundary. Within the Sitka region, all events since 1927 show slip in thedirection of plate motion. To the north in the Cross Sound/Fairweather region the larger eventsshow slip in the direction of plate motion, but the smaller events (M<=6.0) show a clockwiserotation of slip direction relative to plate motion. In the St. Elias, Pamplona and Gulf of Alaskaregions, about 40% of the earthquakes, regardless of size, show slip that deviated more than15 degrees from plate motion, with no consistent direction of rotation relative to plate motionobserved. In the Prince William Sound and Cook Inlet regions, the mismatch between slipvectors and plate motion increases to 60%. This mismatch does not appear to depend onwhether the events occur above or below the plate interface. Most slip vectors for eventslocated beneath Cook Inlet are rotated to the west (counterclockwise) of plate motion, whileevents east of Cook Inlet show greater variability. Events directly below the plate interface inPrince William Sound show east-west directed slip, while events above the plate interface inthe same region have highly variable slip. These shallow events may reflect transitory stresschanges following the 1964 great Alaskan earthquake.


27

The transition from Cretaceous convergence to Cenozoic strike-slip inHispaniola and eastern Cuba

Grenville Draper

Department of Geology, Florida International University, Miami, FL 33199, USA

The presence of Late Cretaceous to lower Eocene magmatic arc rock suites and HP/LTmetamorphic rocks in Hispaniola and Cuba indicate convergence. This convergence wasinterrupted by an oblique collision with the Bahama platform which caused the transition to latePaleogene and Neogene strike-slip dominated tectonics. This collision, as timed by thedeposition of serpentinite deposits, seems to have begun in very latest Maasthrichtian time inCuba, but is Paleocene in Hispaniola. In (?mid) late Eocene time and possibly later, transpression was accommodated south of thepresent Cordillera Central by the formation of the south-verging Peralta/Ocoa fold and thrustbelt. Offset is uncertain, but Pindell and Barrett, 1990 suggest strike-slip component of 350km or more. By latest Eocene/Oligocene, strike slip motion was transferred to the Hispaniola fault zone.Offset of the Amina-Maimon mylonite belt suggests at least 120km of motion, about 50km ofwhich was involved in the formation of the Tavera basin. This motion may have continuedinto early Miocene time. From late Miocene to Pliocene the locus of strike-slip motion jumped northward again to thefault systems of the Cordillera Septentrional. The exact sequence of faulting and the amount ofoffset are uncertain but seems to have been:1. Rio Grande/ Rio Bajabonico faults; offset scaling relationships suggest approx >10kmeach,2. Septentrional fault; offset of amber deposits suggests 85km ?scaling relationships suggest>30km,3. Camu fault; observed offset 50-60km It is not clear when Hispaniola began to separated from eastern Cuba, but consideration ofthe rate of motion of the Caribbean and North American plates (as given by the rates of openingof the Cayman Trough) it was probably in the Oligocene. At the present time strike-slip motion in Hispaniola is taking place simultaneously inHispaniola: offshore and on the Jacagua fault in the Cibao Valley in the north, and along theEnriquillo-Plantain Garden Fault in the south. In addition, north-south shortening is beingaccomodated by underthrusting in southern Hispaniola. The example of Hispaniola suggests that the localization and partitioning of deformation inthe wake of the transition from convergence to strike-slip does not follow any simple geometricor temporal pattern. Pre-existing structures and the geometry of the plate boundary areprobably the most important controlling factors.


28

A Sedimentologic Signature in the Cibao Basin's Neogene Strata of ConvergentStrike-Slip Motion on the Septentrional Fault, northern Hispaniola

Johan P. Erikson

Dept. of Earth Sciences, Dartmouth College, Hanover, NH 03755

Introduction A sedimentologic signature of basement tectonism is identified within the upper Neogene Yaque Group,which accumulated on the asymmetrically subsiding basement of the Cibao basin, northern Hispaniola.Loading of the Cibao basement by south-vergent oblique thrusting along the Septentrional fault zone of theCordillera Septentrional has been established through seismic, gravity studies, and structural studies. Thesedimentary feature to be discussed is an incursion of conglomeratic sediment gravity flow deposits within anotherwise fine-grained section of calcareous siltstone at the base of the Mao Formation on the distal (i.e.,southern) side of the Cibao basin. A critical aspect is that water depths increased at the time of conglomerateinflux, which contrasts with the more common situation elsewhere of a water-depth decrease associated withprogradation of coarse strandline and shoreface facies. Background problem The rates of relative sea-level (RSL) change and sediment supply control shoreline position andaccommodation space within nearshore marine depositional systems. A high rate of RSL rise and a lowsediment supply rate lead to transgression, while a low rate of RSL rise and high sediment supply lead toregression. In shelf and slope settings, transgression is typically interpreted as leading to the generation of thinsections of fine-grained sediment that accumulate over a long time period (cf. Transgressive Systems Tract).This model works just fine for passive margins and similar low-gradient settings with low rates of RSL rise andtransgression. However, many active plate boundaries (convergent or strike-slip) have nearshore depositionalsettings with high-gradient depositional surfaces and occasionally rapid rates of RSL rise and transgression dueto faulting. Based on sedimentary sections of the southern Cibao basin, I am working with a model in whichisolated intervals of coarse sediment gravity flows break an otherwise continuously deepening and fine-grainedshallow-marine sequence, and are the result of a rapid acceleration of transgression and RSL rise. Transgressionand RSL rise accelerate due to contraction across the main, basin-bounding fault, the Septentrional fault.Southern Cibao basin stratigraphy: the Yaque Group. The Yaque Group comprises the vast majority of Neogene strata in the western Cibao Valley. The UpperMiocene to lower Middle Pliocene Yaque Group consists of the Cercado, Gurabo, and Mao formations. Thefollowing is a brief summary of my observations and interpretations of the southern Cibao basin (Erikson et al.1998). Sedimentologic and faunal evidence indicates that the entire Yaque Group preserves one single, continuousepisode of transgression, deepening of the depositional surface, and subsidence. The notable break at the base ofthe Mao Formation in the generally upward-fining trend is not associated with a shallowing event. (Step 1) Transgressive shoreline evolution was dominated by micro- to meso-tidal, wave dominated processesabove a narrow, steep shelf or ramp. Brackish-water and shallow-marine deposits of the Cercado Formation weresucceeded by calcareous siltstones and sandstones with a minor component of Central Cordillera-derived epiclastswithin the Gurabo Formation. The depositional surface was continuously deepening during Cercado-Gurabotime, as inferred from sedimentary structures and the faunal assemblage. (Step 2) The pebble- and cobble-conglomerate comprising a single, 10 to 20 m thick sequence of upward-fining and -thinning sediment gravity flow deposits at the base of the Mao Formation abruptly interrupt anotherwise fine-grained sequence. This occurred as water depths increased (based on sedimentary structures andfaunal assemblages). Subsidence of the Cibao basement must have accelerated during deposition of theuppermost Gurabo and lower Mao Formations. Rapid subsidence was due to uplift of the CordilleraSeptentrional along the Septentrional fault zone.


29

(Step 3) The enlarged sediment supply and possibly steepened depositional slope combined with normal stormactivity and increased seismicity to generate a margin-wide pulse of coarse-grained, sediment gravity flows.Repeated flow events containing coarse, Central Cordillera-derived igneous and Tabera Group clasts createdrepetitive, stacked sequences of generally upward-fining and thinning conglomerates and sandstones at the base ofthe Mao Formation. (Step 4) The supply of coarse clasts decreased after the initial influx, as seen in outcrops. Sedimentationreturned to the pattern established during Gurabo deposition of fine-grained, deep-water, calcareous silt depositioncontemporaneous with localized shallow-water carbonate buildups. Basement subsidence slowed and possiblyreversed in middle and upper Mao-time as indicated by the deposition of the prograding shallow water facieswithin the clays of the upper Mao Formation.Model: Episodic basinal influx of coarse, epiclastic shoreline sediments driven by accelerated transgression (Step 1) A wave-dominated shoreline is transgressing above a moderately steep depositional surface due torelative sea level rise in excess of sediment supply. Much of the epiclastic sediments are retained landward ofthe offshore-transition zone by washover and normal intertidal processes. Those sediments that episodically passonto the shelf and/or slope are deposited in upward-fining and thinning sequences possibly locally affected byreworking storm and geostrophic currents. (Step 2) Acceleration of relative sea level rise occurs due to tectonic or eustatic causes. Transgressionaccelerates because sediment supply cannot immediately increase and compensate for the increased rate of relativesea level rise. Accelerated transgression, however accomplished, leaves a large volume of unconsolidated,coarse, upland-derived sediments below the shoreface zone. The faster the transgression, the greater the amountof the foreshore-backshore prism stranded on the shelf seaward of the shoreface. (Step 3) Storms, seismic events, and possibly a steepened depositional surface (in the case of acceleratedasymmetric basin subsidence) induce sediment gravity flows of the coarse, unconsolidated sediments,transporting them down the depositional slope under high energy conditions. Repeated sediment gravity flowevents lead to deposition of repetitive sequences of graded conglomerates and sandstones. More sediment isproduced during accelerating transgression than during steady-state because transgression is acting more quicklythan sediment supply. The excess sediment is the mobile, unconsolidated sediment above storm wave basewhich was released from the strandline prism during acceleration of transgression. The stratigraphic record, asseen offshore, preserves coarser and thicker sediment gravity flows than were preserved prior to acceleration oftransgression. (Step 4) The rates of transgression and RSL rise stabilize. A constant rate of transgression leads to anequilibrium in which bars/beaches grow to a stable size. As this equilibrium is approached, the volume ofsediments passing offshore is reduced until a steady state is reestablished. Storms and seismic events triggersediment gravity flows which reduce the volume of the excess sediment until equilibrium is reestablished. Oncethe oversupply of relict sediment is exhausted there is only the smaller, background input of sediments that aretransported beyond the intertidal zone, and sedimentation again becomes dominated by finer grained, probablyhemipelagic, sediments. Relevance to the Transitions from Subduction to Strike-Slip Plate Boundaries Formation of the sedimentary feature addressed in this study is initiated by a very rapid RSL rise, which ismost likely to occur due to rapid loading or extension of a basin's basement. The margin-wide incursion ofconglomerate within fine-grained forereef slope deposits (or similar distal accumulations) is a feature that mayfunction as an indicator of rapid basement subsidence. Correct interpretation of water depth changes is critical,because in the absence of reliable water-depth indicators, many relatively coarse-grained intervals are somewhatreasonably correlated with a decrease of water depth. However, for the model described here, an increase of waterdepth is required. One clue to the distinction between these contradictory modes of generation is the clearpresence of a progradational, Highstand Systems Tract (HST) associated with a RSL fall or stillstand, versus theabsence of all but the thinnest HST associated with a rapid RSL rise. To distinguish the proposed model fromtypical lateral migration of coarse, channelized debris flows, there must be multiple exposures along the paleo-shelf margin from which one may be able to deduce that channel migration has not occurred. If the rock recordis reevaluated with an awareness of the possible tectonic significance of margin-wide, anomalously coarseintervals, then more sedimentary evidence of tectonic activity in subduction to strike-slip transitions may befound.


30

Erikson, J. P., J. L. Pindell, G. Karner, L. Sonder, L. Dent, and E. Fuller, 1998, Neogene sedimentation andtectonics in the Cibao basin and northern Hispaniola: An example of basin evolution near a strike-slip-dominated plate boundary: Journal of Geology, v. 106, p. 473-494.


31

Seismic Characteristics of Subduction/Strike-slip Transition Zones asDetermined from Global Earthquake Catalogs

Cliff Frohlich1, Andria Bilich1,2, and Paul Mann1

1Institute for Geophysics, University of Texas at Austin, 4412 Spicewood Springs #600, Austin, TX 78759

2Department of Geological Sciences, University of Texas at Austin, Austin, TX 78713

Earthquake catalogs provide an important source of information about the nature of tectonicactivity along subduction/strike-slip transition zones. We present information on these zonesselected from three global catalogs:

• The Harvard CMT catalog includes focal mechanisms from more than 10,000 earthquakesoccurring between 1977 and 1998. Application of reliability criteria proposed by Frohlichand Davis (1999) eliminates about half of the mechanisms. The remaining mechanismsprovide the principal data for this research.

• The EHB catalog presents carefully relocated hypocenters (Engdahl et al., 1998) for100,000 earthquakes occurring between 1964 and 1995. Application of reliabilitystandards for the number N and distribution (gap) of locating stations (typically, N ≥ 20;gap ≥ 200°; residual ≤ 2 sec) eliminates about a third of the hypocenters. The remainderprovide the principal information about the geographic location of earthquakes used in thisstudy.

• The Abe historical catalog presents locations for about 1500 large earthquakes occurringbetween 1898 and 1976 (Abe, 1981). It is globally complete for earthquakes withmagnitudes of 7 and above. Because infrequent, large earthquakes are responsible formost of the seismically-expressed plate motion, the Abe and CMT catalogs provide thebest available information about geologically significant earthquakes over the past century.

A geographic region meets our definition for a subduction/strike-slip transition zones if itpossesses:

1) a plate-boundary segment with numerous thrust CMT mechanisms (P-axis dip ≥ 50°);2) either a plate-boundary segment with numerous strike-slip CMT mechanisms (B-axis dip

≥ 60°) or a documented strike-slip fault; and3) intermediate-depth (h ≥ 100 km) earthquakes in the EHB catalog situated within the zone.

Applying these criteria, there are approximately 26 qualifying geographic regions. Of these, 7occur at triple junctions. The remaining 19 are regions of progressive transition along acontinuous plate boundary; as explained in the previous presentation, we tentatively classify 14as 'open' and 5 as 'closed' according to their geometery. Several features of the seismicity in these 26 transition zones are remarkable:

• At the 19 zones of progressive transition, earthquakes with strike-slip mechanisms areabsent or relatively rare, even in areas such as the western Aleutians where there are fewthrust earthquakes and intermediate-focus earthquakes disappear. Cumulatively, at thetransition the strike-slip CMTs are far less numerous and than the thrust CMTs;furthermore, taken as a group, the thrust CMTs generally contribute nearly all of the totalmoment (usually > 90%).

• If we compare the record of historical earthquakes to the CMT catalog, an importantquestion is whether this moment deficit is taken up by very large but very rare strike-slipearthquakes, or instead by aseismic motion. Answering this question has significantimplications for evaluating seismic hazard in densely populated transitions zones,including the eastern Caribbean and the Philippines.


32

• At the 7 triple junction transitions zones, strike-slip earthquakes are often numerous,especially if one triple-junction arm is a transform fault in oceanic crust. A typical exampleis the Antarctic-Nazca-South America triple junction in Chile.

• If we classify all CMT into four categories–thrust (P-axis dip ≥ 50°), strike-slip (B-axisdip ≥ 60°), normal (T-axis dip ≥ 60°), and 'other' (satisfying none of the abovecharacteristics)–we find that CMT in the 'other' category are more common in thetransition areas than along regular subduction or strike-slip plate boundaries (see figurebelow).

• In each transition zone we display the relative frequencies of different earthquake typesusing a ternary diagram with vertices corresponding to thrust, strike-slip, and normalmechanism (Frohlich, 1996). Often, the more complex patterns of focal mechanismsseem to occur where the a plate boundary encounters fragments of arc or continental crust.The complexity caused by these fragments complicates the classification of transitionalboundaries. At present, we have been unable to identify any obvious differences betweenthe patterns of focal mechanisms at 'open' or 'closed' boundaries.

There are mechanical reasons why 'other' mechanisms tend to be rare among shallowearthquakes in non-transitional tectonic environments. In particular, the stress conditions thathold at the Earth's surface produce strain releases which correspond to pure thrust, pure strike-slip, or pure normal focal mechanisms. Thus the presence of 'other'-type mechanism isnotable. In some cases it may be indicative of stress conditions which are changing abruptlyalong a plate boundary; alternatively, it may simply reflect the complications introduced by thepresence of arc or continental crust.

Abe, K. Magnitudes of large shallow earthquakes from 1904 to 1980, Phys. Earth Planet. Int.27, 72-92, 1981.

Engdahl, E. R., R. van der Hilst and R. Buland. Global teleseismic earthquake relocation withimproved travel times and procedures for depth determination, Bull. Seis. Soc. Amer., 88,722-743, 1998.

Frohlich, C. Cliff's nodes concerning plotting nodal lines for P, Sh, and Sv, Seis. Res. Lett.,67, 16-24, 1996.

Frohlich, C. and S. D. Davis. How well-constrained are well-constrained T, B, and P axes inmoment tensor catalogs, J. Geophys. Res., (in press) 1999.

Ternary diagram showing distribution of shallow-focus CMT in the subduction/strike-slip transition zone in thewestern Aleutians (left diagram: 17 quakes, 165.5°E-173°E) and in the ordinary subduction region of the centralAleutians (right: 117 quakes, 178°E-174°W). Typically, earthquakes are less numerous and focal mechanism types aremore variable in transition zones than in adjacent regions.


33

Rhyolitic arc-related volcanic corridors at subduction to strike-slip transitions:comparing North Wales and North Island, New Zealand.

Wes Gibbons

Department of Earth Sciences. Cardiff University. Cardiff CF1 3YE, Wales, UK.e-mail: [email protected] .

Some of the most voluminous eruptions of rhyolitic pyroclastics on Earth have occurred from linear clustersof calderas confined within fault-controlled corridors where an oceanic intra- or backarc basin impinges uponcontinental crust. The most spectacular modern example is the Taupo volcanic zone of North Island, NewZealand where over 10,000km3 of ignimbritic ash flow tuff has been erupted over the last 1.6Ma. A similar, butmuch older, example of an arc-related rhyolitic volcanic corridor is partly preserved within late Ordovician rocksin the Snowdonian mountains of northwest Wales, UK. The mid-Caradoc volcanic rocks of Snowdonia, produced by the subduction of Iapetus oceanic crust beneathAvalonia, provide a closely studied ancient analogue for modern arc-related rhyolitic magmatism. They representthe climax of some 35Ma of Ordovician (late Tremadoc to mid-Caradoc) subduction-related magmatism within abroadly elongate (c.120 x >200km) depocenter (the Welsh Basin). Snowdonian volcanicity was restricted in bothspace and time: eruptions took place over just a few (?2-4) million years inside the span of three mid-Caradocstages (Soudleyan, Longvillian, Woolstonian) and two graptolite zones (Dicranograptus clingani andDiplograptus multidens) The tectonic setting invoked for the Welsh Basin at this time was one of subductionand sinistral transtension within an arc founded on continental crust. Snowdonia volcanism was restricted to a NE-SW corridor, now c.20km wide (after Caledonian shortening),within which at least six major rhyolitic calderas, two andesitic stratovolcanoes, and several basaltic ventsdeveloped. Thick (locally >500m) rhyolitic ash flow tuff sequences were erupted from the calderas into a shallowmarine and locally subaerial environment. These rhyolitic products are sometimes intercalated with, and werecommonly followed by, basaltic effusions leaking up bounding faults and constructing transient Stromboliancones and trough-accumulations of submarine lavas. Significant andesite volcanism was confined to thenortheast and southwest ends of the presently exposed corridor, although these eruptions were not exactly coevalbecause eruptions began later in the southwest. Abrupt cessation of magmatism occurred in Woolstonian timesand black graptolitic mud covered the subsiding, extinct volcanic centres. New Zealand currently lies on a transition from oceanic subduction to continental dextral transpression at theAustralian-Pacific plate boundary. The two plates are both spreading away from Antarctica, but variations in theanticlockwise rotation of the Pacific plate has controlled the obliquity of its convergence with the Australianplate. Cenozoic subduction-related magmatism was initiated in earliest Miocene times in the north of NorthIsland where andesites with associated basalts and minor rhyolites were erupted. Miocene arc volcanicity thenfocussed on the Coromandel Peninsula, a north-south belt (c. 35 x 200km) which terminates against theyounger Taupo belt to the south. Early basalt-andesite eruptions in the Coromandel zone were later, in Pliocenetimes, joined by the first major rhyolitic ignimbrite eruptions. More recently, volcanism has reorganised intothe NE-SW oriented Taupo volcanic zone the central part of which has become overwhelmingly dominated byQuaternary rhyolitic ignimbrite eruptions from a cluster of at least 8 caldera complexes. The oldest volcanic centre (Mangakino Caldera) began erupting ash flow tuff around 1.62Ma but now lieswest of the main focus of rhyolitic volcanism which over the last 0.64Ma has been the most productive site ofrhyolitic volcanism in the world. This central zone today forms an actively extending (5-20mm/y) land corridor(120km by 40km) with an extremely high heat flow (c. 2600MW/100 km), and is filled with some 3km ofmostly rhyolitic pyroclastics. Major accumulations of rhyolite in continental crust require wholesale crustal or extreme fractionalcrystallisation of large quantities of ponded, mantle-derived (but preferably crustal contaminated) magma, orsome combination of both. Although crustal fusion has been invoked to explain the Snowdonian rhyolites,most previous workers have favoured basalt fractionation as the major process at work. A similar petrogenetic


34

debate has occurred over central Taupo where rhyolite is hugely dominant over basalts and intermediatefractionates: 95% rhyolite (12,000 km3) as compared to 0.1% basalt (12 km3). This fact, combined with anextraordinarily high heat flow, has led some authors to insist that the bulk of the Taupo rhyolites can only be aresult of significant crustal melting. Central Taupo lies at the southern end of the recently extinct Coromandel volcanic corridor, and heat flow islikely to have been elevated since at least Pliocene times. Quaternary magmatism in central Taupo thereforecaptured pre-heated crustal rocks, and perhaps also reservoirs of mantle melts produced by late Neogenesubduction. Given the density contrast at the Moho, mantle-derived arc-related magma would tend to pond, aneffect further promoted by the increased fracture toughness of a hot, partially molten crust. The petrogeneticpathway required to produce the central Taupo rhyolites is therefore considered to be one involving density-boundunderplating beneath crust with an inherited high heat flow. Melting was perhaps aided by heat generated fromintracrustal plastic deformation which would contribute synergistically to crustal anatexis which in turn wouldinhibit upward propagation of feeder dyke systems and thus reduce transmission to the surface of mantle-derivedmagma. Despite the obvious differences of subduction direction, age, and sense of transcurrent movement, there remainseveral striking similarities in the tectonic settings of Taupo and Snowdonia. In both cases early volcanism was“normal” arc in character, and was heralded by a deformation phase associated with the onset of subduction(Tremadoc in Wales, Miocene in North Island). The first major rhyolitic tuffs were erupted after some 15-20Ma,and the latest phase of volcanism in New Zealand (latest Pliocene-present day), and the final phase in Wales(Soudleyan-Woolstonian) produced the greatest amounts of rhyolitic ash-flow. The Hikurangi forearc has rotated from a WNW trend in the early Miocene to its present NE-SW orientation.During this rotation suprasubduction rollback-induced extension has propagated southwards for 2000km from theoceanic Lau Basin to the Havre Trough, and rhyolitic volcanism has transferred from the Coromandel zone tocentral Taupo. Initial spearheading of the North Island continental crust by the southward propagatingsuprasubduction extension produced the Neogene arc magmatism of North Island. Rollback then rotated the arcsuch that magmatism initially focussed within the Coromandel corridor but then became extinguished ascontinental margin-parallel shear was favoured over subduction. Continued rollback eventually resulted in a moreorthogonal interaction between the axis of arc extension and the continental crust of North Island, favouring areturn to arc magmatism accompanied by suprasubduction extension of the continental crust beneath Taupo.Continued rollback of the Hikurangi subducting slab is promoting clockwise rotations (measured at 7o/Ma insome areas) and extension of the Taupo corridor. It is suggested that the Snowdonian rhyolitic corridor wassimilarly produced by this combination of “spearheading” and rollback-induced rotation of an actively extendingarc basin into Welsh continental crust. There is not a simple continuum between the axes of the Havre Trough and the Taupo zone, but instead theyshow offset along a structure associated with (or at least coincident with) the Vening Meinesz fracture. Analoguemodelling experiments illustrate how oblique impingement of the arc system on the continental margin, whichis weaker than both oceanic and continental crust, could induce shear parallel to the margin and thus explainingthe observed 50km offset along the VMFZ. The apparently sudden extinction of Snowdonian volcanism in mid- to late Caradoc (Woolstonian) times iscompatible with a model whereby magmatism is extinguished as subduction transfers to transcurrentmovements. There is no evidence for collision having terminated subduction beneath Wales: regionalcompression of the Welsh area occurred over 50Ma later, during the Acadian phase of the Caledonian Orogeny.A projection of New Zealand geology 10 million years into the future illustrates how such a transition fromsubduction to strike-slip tectonics might take place. The Hikurangi subduction system, which currentlyterminates gradationally southwestwards into the transpressive Alpine fault system, is predicted to continuerolling back. If this promotes the eastward propagation of the present Alpine fault system and wholesale rotationof eastern North Island, then a future “Hawke Bay” displaced terrane will be dispersed along the Pacific-Australian plate margin. Another consequence of such movements would be the extinction of magmatismwithin the Taupo corridor which will be left stranded behind the future active subduction zone. What likelyhappened in Snowdonia 450Ma ago is thus predicted to repeat itself in North Island, New Zealand in thegeologically near future.


35

Right-lateral Slip along the Southern Boundary of the CaribbeanPlate, Northern Venezuela.

GIRALDO Carlos *, and Franck AUDEMARD

PDVSA Exploracion y Produccion, and FUNVISIS, Caracas, Venezuela.

Several geodynamic models proposed for the Caribbean region imply large-over a 1,000 km long right-lateral displacements along the southern margin ofthe Caribbean plate. In Northern Venezuela, such slip should beaccommodated by the active Boconó, San Sebastián, El Pilar and Oca-Ancónfaults. Nevertheless, this hypothesis seems hardly sustainable when thecumulative slip along every single of these major right-lateral strike-slip faultson Venezuelan onshore is assessed. Such amount of right-lateraldisplacement, as their geodynamic significance, is still matter of controversyin the geological community. From both the revision of the existing literature and our ownexperience,this maximum right lateral displacement is much less than 150 kmsince the individual slip on these major faults are: (a) 30 to 35 km for theBoconó fault, (b) less than 65 km and probably close to 35 km for the Oca-Ancón fault system , and (c) around 60 km for the San Sebastián-El Pilarfault system. The Oca-Ancón and Boconó faults converge nearby Morón andprolong eastward as a single east-west striking right-lateral fault (SanSebastián-El Pilar fault), where kinematics of the former two faults istransferred on the latter one. A general agreement regarding the activation of these major strike-slip faultsseems to exist. A major middle Miocene-Pliocene compressional tectonicregime is responsible for their activation, However, such process has beenslightly dyachronic as it started earlier on the west.


36

Oblique collision of the Bahama Platform in the Hispaniola-Puerto Rico area,northeastern Caribbean

N.R. Grindlay1, P. Mann2 and J. Dolan3

1Department of Earth Sciences, University of North Carolina at Wilmington, 601 S. College Road,Wilmington, NC 284032Institute for Geophysics, University of Texas at Austin, 4412 Spicewood Springs Road, Bldg. 600, AustinTX, 787593Department of Earth Sciences, University of Southern California, University Park, Los Angeles, CA 90089-0740

The 3000 km-long North America-Caribbean (NOAM-CARB) plate boundary that stretchesfrom Guatemala to the Lesser Antilles arc is dominantly a left-lateral strike-slip boundary thataccommodates slow (1-2.6 cm/yr) eastward motion of the Caribbean plate relative to NorthAmerica. The Puerto Rico-Hispaniola segment of the plate boundary is complicated by thepresence of both strike-slip and subduction-related tectonics and the interaction of thesestructures with the obliquely colliding Bahama Platform. The main strike-slip feature is theSeptentrional fault zone located in the Dominican Republic. Fault trenching studies of this faultsegment in the Dominican Republic by Prentice et al., (1993) have shown that its last strike-slip, ground-breaking rupture occurred approximately 700 years ago. The main subduction-related feature is the Puerto Rico Trench located 65-70 km north of the Septentrional fault zone.The Puerto Rico Trench is the surface trace of a south-dipping Benioff Zone that most recentlyruptured during the 1946 M8.0 northeast Hispaniola earthquake (Dolan and Wald, 1997;1998). The Puerto Rico-Hispaniola segment of the NOAM-CARIB plate boundary allows us toaddress several unanswered questions about the effects of obliquely colliding bathymetrichighs on island arcs in strike-slip to subduction settings. In the case of collision, severalquestions arise: how does the collision affect the partitioning of strain along the plateboundary? What are the regional effects of tectonic erosion as the plateau is subducted beneaththe arc? Tectonic escape is a well documented process in continental settings (Molnar andTapponnier, 1975) but how important is this in intraoceanic arc systems? To better understand the complex relation between strike-slip and subduction tectonics northof Puerto Rico, HMR1 bathymetric and backscatter data along with Hydrosweep bathymetryand single-channel seismic, gravity and magnetic data were acquired in June 1996 over thePuerto Rico Trench and northern Puerto Rico/Virgin Island margins (Grindlay et al., 1997).The major features observed include: 1) outer rise of the obliquely subducting Cretaceous-ageAtlantic oceanic crust of the North America plate; 2) the 7-8.4-km-deep trace of this obliquesubduction zone at the Puerto Rico Trench and the flanking North Puerto Rico Slope strike-slipfault, previously identified by Masson and Scanlon (1991); 3) a submerged “forearc” wherewe identified for the first time the South Puerto Rico Slope strike-slip fault (SPRSFZ) strikingparallel to the GPS-derived plate vector (Dixon et al., 1998) and sub-parallel to the NorthPuerto Rico Slope fault 40-60 km to the north; 4) the northern Puerto Rico platform/slope areaoccupied by a drowned Oligocene to lower Pliocene carbonate platform, and 5) the MonaPassage, where we mapped in greater detail the Mona Rift and identified the Yuma Rift for thefirst time. The new sidescan, bathymetric and single-channel seismic data indicate that two anomaloushighs in the northern Puerto Rico and Hispaniola slopes are restraining bends formed betweenthe sub-parallel North and South Puerto Rico Slope fault zones. Both restraining bends appearto be associated with the subduction of high standing ridges on the North American plate. The


37

Mona Block restraining bending exhibits more than 4 km of bathymetric relief and is boundedto the south by the Septentrional fault zone. Uplift of the block is attributed to both the bend inthe South Puerto Rico Slope fault zone-Septentrional fault zone and underthrusting of thesoutheastern extension of the Bahama Platform (Dolan and Wald, 1997; 1998). The maingeologic belts in the region: the outer sedimentary accretionary prism, the metamorphic belt thatextends from the north coast of the Dominican Republic to beneath Mona block and thevolcanic arc south of the Septentrional fault zone change orientation at the Mona Block. TheMain Ridge restraining bend exhibits 2 km of bathymetric relief with thrust faults activelyuplifting its southern margin. The Main Ridge is associated with a higher than average amountof crustal seismicity and is most likely associated with the underthrushing of a fracture zonehigh on the North American plate (McCann and Sykes, 1985; Muszala et al., in review). The survey also revealed extensive normal faulting of the Oligocene to early Pliocenecarbonate platform in the Mona Passage. Two fault populations are observed: 1) arcuate faultsimmediately south of the Mona Block, a pattern interpreted to be related to accentuated archingof the platform due to the collision and underthrusting of the southeastern section of theBahama platform, and 2) NW trending faults which in some instances can be traced on land tolineaments in southwestern Puerto Rico. This localized extension of the platform strata isinterpreted to reflect the recent differential northeastward relative motions of the PuertoRico/Virgin Island and Hispaniola microplates. On the basis of these findings we suggest that the Bahama platform is an “indenter” (Molnarand Tapponnier, 1975) which is presently colliding with the Greater Antilles intraoceanic arcalong the north coast of Hispaniola. This collision has resulted in the change in trend of majorstrike-slip faults and development restraining bends in the forearc region. Anotherconsequence of this collision is the “tectonic escape” of Puerto Rico/Virgin islands block to theENE along the parallel trending strike-slip SPRSFZ and Anegada Fault Zones.

Dixon, T., F. Farina, C. DeMets, P. Jansma, P. Mann and E. Calais, 1998, Relative motion between theCaribbean and North American plates and related boundary zone deformation from a decade of GPSobservations: J. Geophys. Res., 103, 15,157-15,182.

Dolan, J., and D. Wald, 1998, The 1943-1953 north-central Caribbean earthquakes: Active tectonic setting,seismic hazards, and implications for Caribbean-North America plate motions, in J. Dolan and P. Manneds, Active Strike-slip and Collisional Tectonics of the Northern Caribbean Plate Boundary Zone, GSASpecial Paper 326, 143-170.

Dolan, J. and D. Wald, 1997, Comment on “The 1946 Hispaniola earthquake and the tectonics of the NorthAmerica-Caribbean plate boundary zone, northeastern Hispaniola,” by R.M. Russo and A. Villasenor, J.Geophys. Res., 102, 785-792.

Grindlay, N. R., P. Mann and J. Dolan, 1997, Researchers investigate submarine faults north of Puerto Rico,Eos, Trans. AGU, 78, 404.

Masson, D. and K. Scanlon, 1991, The neotectonic setting of Puerto Rico, GSA bull., 103, 144-154.McCann, W.R. and L. Sykes, 1984, Subduction of aseismic ridges beneath the Caribbean plate: Implications

for the tectonics and seismic potential of the Northeastern Caribbean, J. Geophys. Res., 89, 4493-4519.Molnar and Tapponnier, 1975, Cenozoic tectonics of Asia, effects of a continental collision, Science, 189, 419-

426.Muszala, S., N. R. Grindlay, and R. T. Bird, Three Dimensional Euler deconvolution, and tectonic

interpretation of marine magnetic anomaly data in the Puerto Rico Trench, J. Geophys. Res., in review.Prentice, C., P. Mann, F.W. Taylor, G. Burr, and S. Valastro, Jr., 1993, Paleoseismicity of the North

American-Caribbean Plate boundary (Septentrional Fault), Dominican Republic, Geology, 21, 49-52.


38

SUMMARY OF THE DOMINICAN DISASTER MITIGATION COMMITTEE´SACTIVITIES AND IMPACT: 1994-1998.

Christine M. Herridge, Coordinator of Dominican Disaster Mitigation Committee

The Dominican Disaster Mitigation Committee is a solid contribution of the Caribbean Disaster MitigationProject to vulnerability reduction activities in the Dominican Republic. In late 1993, the CDMP beganactivities in the D.R., allowing local businesses and NGO´s the freedom to adapt project objectives to thecountry´s needs and circumstances. As of late 1995, the Dominican Disaster Mitigation Committee, became alegally established NGO to ensure the mid and long term continuation of the CDMP´s vulnerability reductionactivities. The DDMC´s Board of Directors includes 5 NGO´s: Food for the Hungry, World Vision,SODOSISMICA, ADOZONA and ASONAHORES, as well as 4 companies: Codetel, Sea Land Service,Compañía Nacional de Seguros, and J. A. Abreu & Asociados. An independent office was established, equippedand organized to permit efficient and effective access to project-related information, materials, and activities. Thepoor response to products and services promotion, as well as the afiliation campaign, in general, has led to thedesign phase of an ADMD loss estimate for members based on their actual level of preparedness. The ADMDwill develop and market this service and would appreciate any assistance and advice that the CDMP can provide.The ADMD has actively sought cooperative agreements with various agencies and submits proposals to nationaland international agencies interested in supporting vulnerability reduction activities. Already, the ADMD hasbeen awarded with a three year contract with DIPECHO, the first year´s budget will be ECU$100,000 and shouldbegin in December. An equal or greater value contract is needed to complete budgeting for activities andadditional personnel in order to have coordinators for four of the five main activity streams. As a result ofHurricane Georges´ impact, there is a much greater interest in mitigation and various offers of joint activitiesand projects are being considered with NGO´s such as Plan International, World Vision, OXFAM, EsperanzaInternational, IDAC, Cooperación para la Paz, and CIPROS. We would like to take this opportunity to review the DDMC´s five areas of activity as well as the resultsachieved and the targets for 1999.I Training of Trainers and Disaster Management Training: The DDMC has been responsible for fourteen Disaster Management courses, since 1994, providing anopportunity to 329 professionals and technicians from the private, public and NGO sectors to participate inprimarily the OFDA training series. Although other projects and institutions such as the UNDP DisasterMitigation Project through the Secretary of State of Public Works contributed to the training effort, OFDA,CDMP, and the DDMC have made Disaster Management training in the D.R. possible. Due to the success ofOFDA´s effort to train a Dominican team, the DDMC is able to schedule as many courses as desired due to theexperience and capacity of the local team. Of the eight seminars offered by the DDMC in 1998, four successfulcourses were attended by important private company representatives. The seminars generate a modest income forthe DDMC as well. The DDMC has gathered multisectoral interest and support for the implementation of the UNDP´s APELLproject for the Haina Port area and, hopefully, five other high risk locations. This project seeks to create a localresponse capacity among the industrial, official and community stake holders with support from PAHO/OPS,Red Cross, CEPREMID, Ministry of Public Health and other official, community and private organizations.The principal activity will consist of training at various levels as well as exercises and simulations. In early1999 the DDMC will seek the commitment of the private sector and submit proposals to various agencies forassistance with this project. The DDMC offered its training experience and capacity to the OFDA/USAID office in Costa Rica to facilitatecourses in Spanish and English-speaking Caribbean countries and/or to support teams of instructors inCaribbean countries. OFDA did sponsor four Haitian instructors´ participation in the DDMC-organizedTraining for Instructor´s Course in 1997. Eight hurricane simulations have been facilitated by the DDMC, reaching 293 representatives of major hotels,free enterprise zones, insurance brokers, businesses and communities. As a result, the DDMC has receivedformal requests for further simulations for individual companies and insurance firms. These activities are also


39

self-sustaining and serve to generate some income for the DDMC. Another significant training activity includedthe Probable Maximum Loss Seminar held to provide orientation for insurers and reinsurers regarding exposurein case of disaster. The insurers, in most cases, are currently receiving a PML report for their portfolios fromtheir reinsurers. Another factor influencing the lack of industry follow-up on the topic is that catastrophiccoverage was still widely available at reasonable prices. Regarding the CDMP´s interest in offering assistance to Dominican Insurers to implement a similarmitigation plan for home and business owners for a discount on policy rates such as the United InsuranceCompany program available in Barbados: the sector opinion was that the competition was so close that theprices were already cut as far as possible without regard to catastrophic exposure. However, some insurers dodeny coverage to potential clients that they feel are high risks to catastrophic damage and loss. The region willundoubtedly experience some changes in this regards in response to Hurricanes Georges and Mitch.II Coordination and Communication: Since 1994, a total of 1,400 minutes (868 minutes in 1998 alone) of television time worth approximatelyUS$868,965, and 590 minutes of radio time worth approximately US$12,280 have been donated to promote theDDMC´s activities, information and recommendations. An active television campaign was waged worth wellover US$141,900, by the Compañía Nacional de Seguros to prepare the population for hurricane season. The30 second spot was well promoted in July and August and gave equal credit to the DDMC. Additionally, radiostations nation-wide have donated time to run the DDMC´s radio spot on what to do before, during, and afterhurricanes on a daily basis and at no cost. At least 39 newspaper and magazine articles have been published indonated space worth approximately US$23,251. The DDMC has given 326 presentations to a total of 20,616 people nationwide to Dominican businesses,communities, schools and associations regarding the country´s natural hazards and necessary mitigativeactivities. A total of 124 presentations to 9,515 students and teachers in schools in over 20 cities nationwide isincluded in this awareness campaign. The DDMC received generous support from the Compañía Nacional deSeguros which designed colorful and informative materials (in excess of US$20,000) for students. During 1998,the DDMC gave 7 workshops to prepare 140 teachers and 26 Peace Corps volunteers to give the disasterpreparedness and mitigation presentation to students and communities. Presentations have also been given in Miami, Montserrat, Venezuela, and Panama to share the DDMC´ssuccessful experiences. In addition, work meetings are held with heads of businesses, community andgovernment organizations to raise awareness, garner support, obtain information and establish contacts tofacilitate DDMC activities. The DDMC coordinated with the NGO´s involved in its programming and with various international agenciesin preparation for the passage of Hurricane Georges on September 22, 1998. The main focus was on facilitatinginformation regarding damage and needs to assist coordination efforts and the distribution of aid. For thispurpose, the DDMC participated in coordination meetings with the Red Cross and became an active member ofthe Comité de Emergencias Post-Huracán de Organizaciones de la Sociedad Civil organized by ParticipaciónCiudadana, a USAID sponsored NGO which formed an alliance of more than 40 NGO´s to coordinate reliefefforts regarding food, water, health, clothing and territorial reorganization (zoning). The DDMC emphasizedthe introduction of mitigative measures during repair and reconstruction of housing among other activities. During the passage of Hurricane Hortense in 1996, the DDMC maintained contact with threatenedcommunities and local authorities, providing damage estimates and lists of the items needed to organizationssuch as USAID and OXFAM. Fulfilling a basic goal of the CDMP, the DDMC has structured coordination and communication mechanismsto address the needs of communities, private schools and companies, providing a forum in which they can sharesuccessful experiences, gain access to information, videos, books and other materials on disaster mitigationavailable in the DDMC´s office, as well as plan and carry out preparatory simulations before hurricane seasonbegins.III Information: Over 80,000 brochures on what to do before, during and after hurricanes and earthquakes, as well as 5,0001996 calendars detailing hurricane season, have been published and distributed. Twelve informational bulletins on activities, the D.R.´s natural hazards and the most recommended measuresto implement have been edited and distributed to at least 2,500 businesses, organizations and communities (eachissue). A costal flooding map for the Santo Domingo-Haina area was prepared to show levels of flooding foreach category of hurricane. This map, and others regarding hurricane trajectories, flood-prone areas, the deficient


40

drainage system, geological faults, and other graphics have been included in the information bulletins circulatedto date. The DDMC changed the bulletin´s format so that it could be sent by fax/modem and by e-mail every 3months. This change has brought a very positive response from the readership and has cut the bulletin´s costdramatically. Several recent editions have been published and/or features in newspapers and commented on radioprograms. The DDMC has built a video library documenting natural disasters, mitigative activities, project activities andtelevised interviews, DDMC presentations and visits to high risk communities to document the nature of theirvulnerabilities. The video library now holds over 70 VHS tapes, all containing at least 2 hour's worth ofmaterial. The FEMA tapes with hurricane and fire prevention videos as well as the Puerto Rican CivilDefense´s video on earthquakes and the DDMC´s video with the Dominican Red Cross on how to handleaccident victims are constantly in demand. The DDMC holds Annual Meetings each November, and this year the Presentations by the Board Members ofproject activities were accompanied by video clips of the activities. The outstanding contributions made byindividuals and organizations were recognized with plaques.IV Community Education: In 1995, the DDMC evaluated and selected a program designed by the Red Cross and has, to date, sponsored 6courses to train 173 local facilitators from 44 NGO´s, 12 Project Development Areas (PDA´s) sponsored byWorld Vision, as well as from the Civil Defense and the Red Cross. Since October of 1995, 578 high riskcommunities have received the Community Disaster Preparedness Workshop. Over 17,340 community leadershave learned what a disaster is, how to identify vulnerability, how to identify the community's human andmaterial resources, and - as a community - how to design a Community Emergency Plan. More could havebeen accomplished in 1998 if the facilitator training course had been given earlier in the year, which was notpossible due to World Vision´s internal difficulties. Also, the active hurricane season made programmingdifficult and NGO´s fell behind in implementation. In July of 1996, the workshop facilitators attended an evaluation seminar sponsored by the DDMC and theInternational Federation of Red Cross and Red Crescent Societies to discuss problems and solutions regardingthe campaign. As a result, several improvements and changes have been implemented to standardize and assurethe quality of the workshops and reduce the cost per workshop by at least a third. The goal for the 1998campaign is to continue to increase local counterpart contribution, currently at 50% of the program´s value, to90%. Thanks to an agreement with the World Food Program (UN) this significant cost reduction is possible.The DDMC focuses its resources on training facilitators in new regions and supports current participatingNGO´s with miminal materials needed to give the workshops. This transition is already underway throughNGO´s like World Vision which provides an operating budget to 13 NGO´s, each of which is required to dedicatea portion of its resources to disaster mitigation activities. The DDMC hopes to integrate the Dominican Red Cross into its expansion efforts now that Dominican lawprotects its sovereignty as an NGO and the IFRC has moved its headquarters to Santo Domingo. The RedCross is, in many cases, the only NGO with a strong presence in important areas surrounding La Vega,Santiago, Puerto Plata and other areas of the Cibao. Hopefully, past incapacities can be overcome to build upan effective community awareness program within the Red Cross´ sphere of activity.V Community Initiatives: The DDMC´s facilitation of Community Initiatives to implement vulnerability reduction projects wasdesigned in consultation with NGO´s in the DDMC´s Board of Directors. The Community Initiative Facilitatorwas hired in August of 1996 and has actively oriented qualifying communities regarding the C.I. program andthe application process. Given the requirement that the communities must provide - in local counterpartcontributions such as manual labor, technical assistance, land, project permits, etc.- at least 50% of the totalvalue of the project, fewer applications were received than initially anticipated. During the period of August, 1996 through November, 1998, the DDMC has assisted a total of 17 rural andurban communities with various small infrastructure projects to reduce vulnerability regarding flooding andlandslides. The investment made so far amounts to US$281,267.00, of which the local communities providedUS$129,023; and other sponsors have contributed US$63,173; and the DDMC has invested US$89,070, only32% of the entire value of the program. A total of 5,598 families representing 30,789 people have benefitteddirectly from these projects. In addition, 25,652 people from 4,664 families have benefitted indirectly. Thetotal beneficiary population reaches 56,441 people in 10,262 families nationwide.


41

Twelve of the projects are drainage embankments leading to over 3,357 meters of channels to reduce flooding.One is a river containment dike. Two projects are a series of bridges over drainage channels. Another projectbuilt drainage wells and and the remaining project resulted in the construction of two containment walls toprevent landslides. Fifteen of the seventeen projects have been completed. Two should conclude within the firsthalf of December and will be inaugurated in December, along with another that was just completed. Videos andpictures are available regarding each project. In spite of periodic heavy rains since January and the passage of Hurricane Georges, 15 of the 17 projects havesuccessfully fulfilled their objectives to the delight of the communities. Field reports show that the projects hada very significant effect in reducing the flooding expected due to the Hurricane. The two projects which faileddid significantly reduce the impact of the Hurricane even though the failures did occur and have since beenassessed and the modifications in design will be implemented when the reconstruction takes place.

SUBDUCTION, STRIKE-SLIP AND COLLISION IN THE ALPINE-CARPATHIAN REGION

Part 5: Seismicity and Recent stress field of the Eastern Carpathian lithosphere:A Recent snapshot of terminal subduction retreat

Stefan Hettel1, Birgit Müller1, Blanka Sperner1, Friedemann Wenzel1, Karl Fuchs1

1 Geophysical Institute, University of Karlsruhe, Hertzstr. 16, 76187 Karlsruhe, Germany

SeismicityModerate crustal seismicity is recorded all over Romania (max. magnitude 5.6), but a far more

intense subcrustal activity is restricted to a small seismogenic volume in the SE-bend of theCarpathian arc (Fig. 8), with about 30 x 70 km2 lateral and 130 km vertical extent. Thisseismogenic volume correlates with the extent of the high-velocity body detected by seismictomography (part 4). Northeastern and southwestern crustal bounds of the volume are the Trotus-Peceneaga-Camena fault zone and the Intramoesian fault, respectively. This focus of seismicevents represents the most severe seismic hazard in Europe today: every century more than 20earthquakes with magnitudes greater than 6.0 are recorded (max. magnitude 7.7 in 1940).

The moment release of the last seven large events is estimated to 1.0·1021 Nm. Averaging overa large time interval and including earthquake reports of the pre-instrumental recording time, thisresults in a moment release rate of 8·1018 Nm a-1. This value is associated with a verticalextension of the Vrancea seimogenic volume at a rate of about 2.0·10-7 a-1.

Fault plane solutionsThe crustal events (max. depth 35 km) show no uniform orientation of stress axes and nodominant tectonic regime. The stress axes orientations of the subcrustal events indicate a clearthrust-faulting regime, with the largest events having nearly identical orientations of fault planes.This has a direct impact on the seismic hazard in the region, because earthquakes with nearlyuniform focal mechanisms should produce a similiar wavefield on the surface and thus constrainthe distribution of the hazard.

Stress inversionStress inversion with all available focal mechanisms from the subcrustal seismogenic volume

gives a best-fit stress-tensor of 079/19 (σ1), 341/21 (σ2), 207/61 (σ3) with a misfit-angle of 22°.Inversion of datasets from depths between 40 and 140 km indicates E-W compression, whereas indepths below 140 km N-S compression dominates. The depth range in which the change indirection takes place correlates with the change in strike direction of the high-velocity body seenby seismic tomography.

Stress fieldBesides earthquake focal mechanisms, crustal stress information can be obtained from

borehole-breakout measurements (in the upper few kilometers) and from the analysis ofgeological structures. In the central Pannonian Basin the observed crustal stress field withmaximum horizontal compression in NE-SW orientation can be related to the continentalcollision in the Dinarides. In contrast, the stress data from the Eastern Carpathians and adjacentforedeep basin show neither a uniform stress direction, nor a dominant tectonic regime (Fig. 9).Possible reason for this complexity is a superposition of different stress signals: stressreorientation along the major faults, pull of subducted slab, local basin formation and far-fieldtectonic stresses.

Fig. 8: Recent recorded earthquakes with magnitude greater than 3.0 in Romania. The moderate crustalactivity (white squares) is more widespread, whereas the intense subcrustal seismicity (grey squares) islocalized in the SE-bend of the Carpathian arc, indicating a subducted slab.

Fig. 9: Orientation of recent maximum horizontal compression in SE-Europe. The symbols are as indicatedin the legend. Gridded black lines show the mean orientation of SH. in the area. Whereas the observed stressfield in the Central Pannonian Basin is uniform, is the observed stress pattern in the Eastern Carpathiansvery complex, due the superposition of various different stress signals.

18

18

20

20

22

22

24

24

26

26

28

28

30

30

44 44

46 46

48 48

0 50 100

km

BelgradeBucharest

Budapest

Method:focal mechanismbreakoutsovercoringhydro. fracturesgeol. indicators

Regime:NF SS TF U

Quality:ABC

depth interval: 0 - 35km


44

KINEMATICS OF WRENCH ZONES

Catherine Hier and Christian Teyssier

Department of Geology and Geophysics, University of Minnesota, Minneapolis, MN 55455, USA.(e-mail: [email protected] and [email protected])

The continental lithosphere is vertically stratified into three layers of contrasting rheology: the brittleupper crust, the ductile mid-lower crust, and the upper mantle. Wrench zones provide an ideal setting to studythe difference in behavior of these rheological layers during deformation because the vertical structure of thelithosphere is not disturbed during deformation as it is along extensional and contractional zones.

The deformation in the brittle upper crust along wrench zones is accommodated by strike-slippartitioning. Rigid blocks are translated along strike-slip faults sub-parallel to the plate boundary. Deformationis concentrated along these strike-slip faults and the crust within the blocks remains essentially undeformed (Fig.1).

Recent developments in seismology have allowed the resolution of mantle fabrics from seismic anisotropydata. This data has provided a glimpse of the structural geology of the upper mantle. Shear-wave splittingstudies along wrench plate-boundaries have shown that a fast polarization direction lies parallel to the directionof wrenching. This zone of fast polarization has been well documented along the San Andreas plate boundary(Özalaybey and Savage, 1995) and both the North American and South American boundaries of the CaribbeanPlate (Russo et al, 1996). The zone of fast polarization is interpreted by Teyssier and Tikoff (1998) as a zone ofpenetrative shearing in the upper mantle along the plate boundary (Fig. 1).

Fig . 1 Conceptual model of a three-layer lithosphere at a wide wrenching plate boundary (Bourne et al., 1998). Uppercrustal blocks are translated above the shearing mantle. The grey zone is the ductile crust that transfers motion frommantle to upper crustal blocks. Arrows are velocity vectors.

The mechanics of the ductile crust sandwiched between the upper crust and upper mantle is not wellunderstood. This layer, however, must accommodate the differential displacements between the strike-slippartitioned motion in the upper crust and the penetrative shearing in the upper mantle.

The kinematics of the ductile crust sandwiched between the rigid blocks of the upper crust and the uppermantle simple shear zone below can be modeled mathematically. This deformation zone combines a verticalshear zone where wrench is maximum at the base and 0 at the top with a horizontal shear zone where shearing ismaximum at the top and 0 at the base. Although an analytical solution for the strain matrix in the ductile crustcannot be determined because deformation is inhomogeneous, solutions can be approached for a series of discretethin layers superposed in the vertical direction. Within each of the thin layers, strain is produced by the twoshear components mentioned above and corresponds to a simple shear for which instantaneous and finite strainquantities can be determined. This type of forward modeling predicts orientation of foliation, lineation, shear


45

plane and direction within each thin layer, and the variation in fabric orientation from top to bottom of thedeformation box. Any component of transpression can also be accounted for mathematically.

An alternative approach rooted in fluid mechanics can also be used to estimate strain within the ductilecrust. If the rigid blocks are translated above a shearing lower lithosphere, and the ductile crust consists of aNewtonian fluid, then the thickness of the coupling zone is related to the width of the individual crustal blocks(Bourne et al. 1998). For example, for blocks 15-20 km wide, the transition zone in the ductile crust iscalculated to be 5-7 km thick, from Laplace's equations. For a non-Newtonian fluid of power-law rheology, thisthickness would decrease because deformation would be more localized. Therefore, it is critical to make geologicobservations of the ductile crust in such settings and study the structures, microstructures, and deformationmechanisms of the deformed rocks in order to place mechanical bounds on rock rheology.

The Northern Range of Trinidad provides a unique opportunity to study an exhumed segment of ductile crustdeveloped along a plate-scale wrench zone. The Northern Range of Trinidad preserves a cross-section from theupper crustal rocks in the eastern end of the range to ductilely deformed mid-crustal rocks at the western end ofthe range (Weber et al., in review).

The mantle shear zone north of Trinidad, documented from seismic anisotropy by Russo et al (1996) alongwith the transition from upright low temperature fabrics to higher grade flat lying fabrics exposed in theNorthern Range of Trinidad preserve a complete vertical section of the lithosphere that was deformed duringwrench motion along the Caribbean/South American Plate Boundary. Detailed studies of the ductile crustpreserved in the Northern Range of Trinidad will allow us to address how the motion of the upper mantle istransferred vertically through the ductile crust into the brittle upper crust.

REFERENCESBourne, S. J.; England, P. C., and Parsons, B. (1998). The motion of crustal blocks driven by flow of the

lower lithosphere and implications for slip rates of continental strike-slip faults. Nature. Vol. 391. pp.655 - 659.

Özalaybey, S. and Savage, M. K. (1995). Shear-wave splitting beneath the western United States in relation toplate tectonics. Journal of Geophysical Research. Vol. 100. pp. 18,135 - 18,149.

Russo, R. M.; Silver, P. G.; Franke, M.; Ambeh, W. B.; and James, D. E. (1996). Shear-wave splitting innortheast Venezuela, Trinidad, and the eastern Caribbean. Physics of the Earth and Planetary Interiors.Vol. 95. pp. 251 - 275.

Teyssier, C. and Tikoff, B. (1998). Strike-slip partitioned transpression of the San Andreas fault systems: alithospheric scale approach. In Continental Transpression and Transtension Tectonics. pp. 143 - 158.

Weber, J. C.; Ferrill, D.; and Roden-Tice, M. (in review). Calcite and quartz microstructural geothermometry oflow-grade metasedimentary rocks, Northern Range, Trinidad. Journal of Structural Geology.


46

Uplifted marine terraces on the east side of Bering basin along the shorelinesof Bering and Karaginskiy islands: Indicators of large earthquakes in theKamchatka/Aleutian region?

Kurbatov A.V.1, Bursik M.I.1, Melekestsev I.V.2, Sulerzhitsky L.D.3

1. Department of Geology, SUNY at Buffalo, 876 NSM Complex, Buffalo NY, USA 142602. Institute of Volcanic Geology and Geochemistry, 9 Piip av., Petropavlovsk-Kamchatsky, Russia 6830063. Geological Institute, 7 Pyizevskiy st., Moscow, Russia 109017

Tephrochronological dating (Braitseva et al., 1997) of co seismic-uplifted marine terraces onthe east side of the Bering Basin allowed us to determine the age and extent of greatearthquakes along the Kamchatka Peninsula for the late Holocene (Fig.1).

The radiocarbon ages of the first layers formed after terraces emerged from the sea followedthe last sea-level transgression. For peat sampled in Karaginskiy Island the age is cal. BC4470(4460,4410)4360 (GIn7778) (all samples were calibrated using Stuiver and Reimer (1993)for one standard deviation); for a peat layer at Uka site the age is cal. BC 5370(5280)5260(GIn7794); and for a wood fragment, cal. BC 5060(5000)4940 (GIn6329), in the “Cherniy Yar”site. The different elevations (up to 10 m) of these layers in relation to present sea level supporta co seismic mechanism of terrace formation.

The time range of terrace formation on Bering Island was determined based on radiocarbonages for mollusc shells (cal. BC 1120 (980) 850 (GIn7750) and cal. BC 750 (510) 550 (GIn7751)and the peat layer above uplifted lake sediments cal. BC 370 (340, 320, 200) 170 (GIn7752).Another episode of uplift took place before deposition of a Shiveluch tephra layer dated by anoverlying peat layer at cal. AD 1020(1070,1090,1120,1140,1150)1210 (IVAN 446). It is unclearwhen the lowest surface, 1.5 m above sea level, was formed because of the absence of any ashlayers, which could be also due to many different factors such as tsunamis, storm-waveerosion or winds.

Two relatively well-preserved surfaces along the shoreline of Ossora settlement were formedbefore the deposition of the Ksudach (Ks1) cal. AD 240 ash on a 4.5 m terrace, and Shiveluchcal. AD 1230(1260)1280 on a 2m surface (Melekestsev et al., 1997).

For Karaginskiy Island, the age of the best-developed 4.5-5 m–high terrace, is older than anash fall from Shiveluch volcano of c. 1600 yr BP correlated based on a radiocarbon date of cal.AD 260(390)424 (GIn 7783) obtained from a peat layer above the tephra. Comparison of fieldobservations made in 1993 with aerial photographs taken before the Ms=7.7 tsunamigenicearthquake on the eastern coast of Ozernoi Peninsula of November 23, 1969 57.8 N, 163.8 Eshows that the channel of the Gnumvayam River changed after massive sand depositionblocked the river mouth, probably during tsunami propagation. A new terrace surface was notfound as a result of this event which supports the hypothesis that only large seismic eventsgenerate uplift.

Clusters of marine terraces along the shoreline of Kamchatskiy cape show at least 4 episodesof co-seismic uplift (Melekestsev et al., 1994): 1) before deposition of the Ks1 tephra layer ofcal. AD 240, 2) before deposition of Shiveluch cal. AD 640(650)660 tephra, 3) before depositionof Shiveluch cal. AD 1230(1260)1280 tephra, and 4)one event before the Shiveluch eruption in1964 AD, probably related to the largest known historical earthquake in November of 1737(Zayakin and Luchinina, 1990).


47

Although more data have to be collected, some preliminary conclusions can be drawn. Thehighest tectonic activity occurred around AD 200 along Kamchatka’s Pacific Pate boundary.

ReferencesBraitseva O.A., Ponomareva V.V., Sulerzhitsky L.D., Melekestsev I.V., Bailey J., 1997,

Holocene key-marker tephra layers in Kamchatka, Russia, Quaternary Research, 47, pp.125-139.

Melekestsev I.V., Kurbatov A.V., Pevzner M.M., Sulerzhitsky L.D. “Prehistoric tsunami andlarge earthquakes in the Kamchatskiy peninsula (Kamchatka) as shown bytephrochronological data, 1994, Vulkanologija i Seismologija, n.4-5, pp.106-115 (inRussian)

Melekestsev I.V., Kurbatov A.V. The frequency of Great paleoearthquakes on thenorthwestern coast of the Bering sea and in Kamchatka Basin During Late Pleistocene/Holocene time, 1997, Vulkanologija i Seismologija, n.3, pp. 3-11.(in Russian)

Stuiver, M. and Reimer, P.J., 1993, Radiocarbon, 35, pp. 215-230.Zayakin Y.A., Luchinina A.A., 1990, Catalog of tsunamis in Kamchatka, Springfield: U.S.

Dept. of Commerce National Technical Information Service, 74p.


48

Extracting Plate Kinematic Information from Mélange Fabrics

Timothy M. Kusky1 and Dwight C. Bradley2

1Center for Remote Sensing, Boston University, Boston MA 02215 USA2U.S. Geological Survey, Anchorage Alaska, 99508 USA

Analysis of early deformational fabrics in tectonic mélange may yield information about the kinematics ofpast plate interactions (e.g., Kano et al., 1991; Onishi and Kimura, 1995; Kusky et al., 1997a; Kusky andBradley, in review). In this paper we develop a methodology for using kinematic data derived from tectonicmélanges in accretionary prisms to derive information about the slip vector directions within an accretionarywedge setting. This information may ultimately prove useful for reconstructing the kinematic history of plateinteractions along ancient plate boundaries, and how this convergence was partitioned into belts of head-on andmargin-parallel slip during oblique subduction, and for obtaining a better understanding of subduction - to strike-slip transitions.

As an illustration of the method, we analyze asymmetric fabrics from mélange of the McHugh Complexfrom the Seldovia Quadrangle on the southern part of the Kenai Peninsula, Alaska, with particular emphasisplaced on relations from specific areas in which glacial retreat has created excellent 3-D exposure, permittingdetailed mapping (Bradley and Kusky, 1991; Bradley et al., 1999; Kusky and Bradley, in review). The McHughComplex is a Mesozoic mélange comprising part of the Chugach terrane of southern Alaska, interpreted as anaccretionary prism formed by offscraping (and/or underplating) outboard of the present seaward margin of thecomposite Peninsular - Wrangellia - Alexander superterrane. It is subdividable into subparallel internallycomplex belts of dominantly Triassic through Cretaceous chert, basalt, gabbro, graywacke, argillite, exoticPermian limestones, and other belts mappable macroscopically only as mélange, all of which wrap arounddismembered ultramafic massifs, and intruded by a variety of igneous rocks (Bradley et al., 1999). We havestudied and mapped the McHugh Complex at scales ranging from 1:250,000 through 1:1,140, to 1:1, andmapped thin sections at magnifications of up to 250:1. A scale-invariance is reflected in the map patternsthrough the thin section scale, with individual belts varying from a fraction of a mm to a few km thick.

Structures within the mélange indicate a history including both contractional and extensional strains, formedat a variety of depths of deformation. A map-scale oceanic plate stratigraphy is repeated hundreds of times acrossthe area, but structures at the outcrop-scale are dominantly extensional. Rheological units that were strongerthan surrounding units during deformation now occur as blocks within a matrix that has flowed around thecompetent blocks, forming broken formation and mélange. Deformation was noncoaxial, and disruption ofprimary layering was a consequence of general strain driven by plate convergence, in a relatively narrow zonebetween the overriding lithified accretionary wedge and the downgoing, thinly-sedimented oceanic plate. Soft-sediment deformation processes did not play a major role in the formation of the mélange. We propose a modelin which layers oriented at low angles to σ1 are contracted in both the brittle and ductile regimes, layers at 30°-45° to σ1 are extended in the brittle regime and contracted in the ductile regime, and layers at angles greater than45° to σ1 are extended in the brittle and ductile regimes.

Many of the structures within mélange of the McHugh Complex are asymmetric, and record kinematicinformation consistent with the inferred structural setting in an accretionary wedge. We develop a method forinterpreting the kinematic significance of the fabric asymmetry in mélange, following in part earlier work byKano et al. (1991). Slip vectors can be derived from many of the structural elements present in mélange, andanalyzed in a manner similar to kinematic data from higher-grade mylonitic rocks. However, rocks in mélangelike the McHugh Complex commonly do not have well developed slip lineations, so derivation of kinematicdata is less direct. We have used the following asymmetric fabric elements for kinematic analysis: 1)intersection of planar elements, such as C (Y), S, P, and R1 surfaces, or between thrusts within duplexstructures, with the slip vector lying at 90° to the intersection lineation, in the plane of slip (C); 2) slip orslickenline lineations on slip surfaces; and 3) fragment elongation lineations and bouden axes within S - surfacesof mesoscale mélange.

The most reliable data are plotted on the map as a series of lower-hemisphere, equal angle projections. Foreach of the stations great circles for C (Y) surfaces are plotted as solid lines, S- and R1 surfaces as dashed lines,and the slip vector direction in present coordinates is plotted as a solid dot with an arrow through it pointing inthe direction of slip of the hangingwall. Because rocks of the McHugh Complex have been rotated sinceformation of the mélange, and folded about vertical axes (e.g., Kusky et al., 1997b), several geometric


49

corrections to the data were made to restore them to their presumed original attitude. First, a rotation about avertical axis was performed on the data to bring the local strike into parallelism with the regional 035° strike.Second, the data were rotated about a horizontal axis to account for tilting and/or folding or the layering in themélange. Since there is considerable uncertainty about the attitude at which the asymmetric fabrics formed, therotated slip vectors are plotted for an attitude restored to a 30° NW dip for the main mélange foliation, and alsoin an attitude in which the slip vectors are restored to horizontal. Our assumptions in such rotations areaccounted for by plotting this range of possible initial orientations for the slip vector orientations of thehangingwall blocks in the mélange.

Mélange of the McHugh Complex on the lower Kenai Peninsula can be divided into three sub-parallel beltswith different slip vector orientations. The most inboard belt shows generally E-W directed slip vectors, withone exception coming from a chert unit that shows complex disharmonic folding on a regional scale (Kusky etal., 1997b; Bradley et al., 1999). A central belt preserves NNE-SSW directed slip vectors, and an outboard beltpreserves SW-NE directed slip vectors. Together, these slip vectors can be interpreted as a displacement fieldassociated with initial formation of the mélange.

The tectonic significance of the displacement field is less clear. In the simplest case, as the mélange growsprogressively by successive offscraping events, each package will record the plate convergence direction at thetime of accretion. By examining progressively younger (more seaward) mélange packages, a history of plateconvergence directions may be reconstructed. In this sense, accretionary mélanges may record a kinematichistory of convergence directions and plate interactions that occurred along a convergent margin duringproduction of the mélange. Kinematic analysis of mélange fabrics may therefore yield kinematic informationthat is complimentary to, and extends the temporal limit of sea floor magnetic anomaly data that is commonlyused for plate reconstructions. Systematic analysis of the kinematics of tectonic mélange belts formed atconvergent margins may yield information about the convergence directions between the overriding andunderriding plates, the mechanics of accretionary wedges, and the partitioning of deformation in zones of plateconvergence. However, some complications in such an interpretation arise since accretionary wedges may notshow such a simple direct response to plate convergence directions. Wedges may extend or contract in responseto body forces and surface slopes (e.g., Platt, 1986), and strain may be partitioned into belts that showpredominantly head - on convergence, and belts that show predominantly orogen-parallel displacements (e.g.,Dolan and Mann, 1998). Further, more detailed kinematic analysis, coupled with well-controlled age-data andpaleomagnetic poles may yield additional insight into the relationships between the kinematics of mélange inaccretionary wedges and plate convergence directions.

Bradley, D.C., and Kusky, T.M., 1992, Deformation history of the McHugh accretionary complex, SeldoviaQuadrangle, south-central Alaska. In Geologic Studies in Alaska by the U.S. Geological Survey during1990, ed. D.C. Bradley, and A. Ford, Bulletin of the U.S. Geol. Survey 1999, 17-32.

Bradley, D.C., Kusky, T.M., Haeussler, P., Karl, S., and Donley, D., 1999, Geologic Map of the SeldoviaQuadrangle, U.S. Geological Survey Open File Report.

Dolan, J.F., and Mann, P., 1998, Active strike-slip and collisional tectonics of the northern Caribbean plateboundary zone, Geological Society of America Special Paper 326, 174 pp.

Kano, K., Nakaji, M., and Takeuchi, S., 1991, Asymmetrical mélange fabrics as possible indicators of theconvergent direction of plates: a case study from the Shimanto Belt of the Akaishi Mountains, central Japan.Tectonophysics 185, 375-388.

Kusky, T.M., and Bradley, D.C., Kinematic Analysis of Mélange Fabrics: Examples and Applications from theMcHugh Complex, Kenai Peninsula, Alaska. Journal of Structural Geology, in review

Kusky, T.M., Bradley, D.C., and Haeussler, P., 1997a, Progressive deformation of the Chugach accretionarycomplex, Alaska, during a Paleoene ridge-trench encounter. J. Struc. Geol. 19, 139-157.

Kusky, T.M., Bradley, D.C., Haeussler, P.J., and Karl, S., 1997b, Controls on accretion of flysch and mélangebelts at convergent margins: Evidence from the Chugach Bay thrust and Iceworm mélange, Chugachaccretionary wedge, Alaska. Tectonics 16, 855-878.

Onishi, C.T., and Kimura, G., 1995, Change in fabric of mélange in the Shimanto Belt, Japan: Change inrelative convergence? Tectonics 14, 1273-1289.

Platt, J.P., 1986, Dynamics of orogenic wedges and the uplift of high-pressure metamorphic rocks. GeologicalSociety of America Bulletin 97, 1037-1053.


50

Strain partitioning and coupling between plates:Field observations and critical input from experimental modeling

Serge Lallemand 1 and Alexander Chemenda 2

1 Laboratoire de Géophysique et Tectonique, CNRS, Université Montpellier 2, CC.60, place E. Bataillon,34095 Montpellier, France – email: [email protected]

2 CNRS, Université de Nice - Sophia Antipolis, 06560 Valbonne, France – email: [email protected]

What are the forcing functions which govern strain partitioning in oblique subduction zones? It can be (1) the convergence obliquity, (2) the strength of the overriding plate, (3) the degreeof interplate coupling, (4) the tectonic regime of the overriding plate, or (5) the slab pull. Oneof the most difficult parameter to test is the interplate coupling, because there is not a singleway to estimate it in nature (seismic coupling ?, kinematic coupling ?, viscous coupling ?,interplate pressure ?).

We have performed 40 experiments of 3-D analog modeling of oblique subduction to studythe mechanism of strain partitioning. The model is two-layer and includes elasto-plasticconvergent lithospheres and a low-viscosity liquid asthenosphere. The subduction is driven bythe push force from a piston and the pull force when the density contrast between thesubducting plate and the asthenosphere is positive. We vary both the slab pull and the interplatefriction (frictional stresses), keeping the angle of obliquity constant and equal to 40°.

(1) Slip partitioning was only observed in the models with high interplate friction andwhen the overriding plate contained a weak zone (cut or thinned lithosphere).

(2) Back-arc extension and transcurrent faulting within the overriding plate appear to benon compatible processes.

The along-strike component of the friction force Fft is balanced with the shear resistancealong the transcurrent strike-slip fault Fss and the boundary forces (driving or resistive) Fc andFe. The horizontal component, normal to the trench, of the friction force Ffh is always negative(compression), whereas the horizontal component of the slab pull Fph is either negative orpositive. The sum of these two horizontal forces Fh governs the tectonic regime within theoverriding plate.


51

Figure 1: Balance of forces in oblique subduction zones. The interplate friction force Ff is the main force whichhas an along-strike component Fft able to drive the forearc sliver. It may exist additional forces at bothextremities of the sliver such as Fc and Fe.

The conclusion (2) is supported by the comparison of two oblique subduction zones whosegeometry is comparable. The southern Kuril arc-trench system is compressional and undergoesstrain partitioning with a convergence obliquity of only 27°, whereas the southern Ryukyuzone, with an obliquity of 50°, is extensional and no transcurrent faulting is observed in thebasement of the arc.

Deflection of slip vectors in the southern Ryukyu subduction zone together with the absenceof transcurrent faulting within the arc basement can be explained by a shear within thesubducting slab. Indeed, strain partitioning may occur either in the overriding or the subductingplate if special conditions are met, such as the presence of a weakness zone within thelithosphere. Lateral forces are expected within the subducting slab in the southern Ryukyus asa consequence of the arc-continent collision occurring in Taiwan.

To summarize, existence of interplate friction and weakness zone in overriding plate arenecessary conditions for strain partitioning in oblique subduction zones. The degree of slippartitioning will depend on the friction force and the boundary forces. The tectonic regimewithin the overriding plate depends on the interplate pressure which is slab pull and frictiondependant. The friction force could be "slab pull" dependant, as suggested by the apparentinhibition of strain partitioning when the overriding plate undergoes extension.


52

Trench-parallel stretching and folding of forearc basins and lateral migrationof the accretionary wedge in the southern Ryukyus: A case of strain partitioncaused by oblique convergence

Serge Lallemand 1, Char-Shine Liu 2, Stéphane Dominguez 1, Yvonne Font 2, PhilippeSchnürle 2 and Jacques Malavieille 1

1 Laboratoire de Géophysique et Tectonique, CNRS, Université Montpellier 2, CC.60, place E. Bataillon, 34095Montpellier, France – email: [email protected]

2 Institute of Oceanography, National Taiwan University, P.O. Box 23-13, Taipei, Taiwan –email:[email protected]

Detailed seafloor mapping in the area east of Taiwan revealed trench-parallel stretching andfolding of the Ryukyu forearc and lateral motion of the accretionary wedge under obliqueconvergence (Fig. 1).

Fig.1: Present-day kinematics of the Taiwan region as given by GPS measurements. Arrows are displacementvectors of structural blocks with respect to the South China Block.


53

East of 122°40'E, a steep accretionary wedge is elongated in an E-W direction. A majortranscurrent right-lateral strike-slip fault accommodates the strain partitioning caused by anoblique convergence of 40°. A spectacular out-of-sequence thrust may be related to thesubduction of a structural high lying in the axis of the N-S trending Gagua Ridge. This asperityis likely responsible for the uplift of the accretionary wedge and forearc basement and mayhave augmented strain partitioning by increasing the coupling between the two plates.Morphological and structural considerations on wedge uplift and reentrant above and in thewake of subducting asperities have allowed us to estimate the azimuth of the relative motionbetween the subducting plate and the accretionary wedge. This azimuth is deflected by 20° withrespect to the convergence between the converging plates as given by GPS measurements.Assuming that all the lateral component of the relative motion is accounted by the transcurrentfault, it gives a slip rate of about 3.7 cm/yr along the fault.

No major transcurrent fault has been mapped within the Ryukyu arc basement or back-arcbasin (southern Okinawa Trough), despite the high convergence obliquity (40 to 60° dependingon the sectors). Furthermore, GPS measurements indicate that the southern Ryukyu Arc isdrifting toward the south-southeast with respect to the South China Block (including Chinesecontinental shelf), rather than toward the southwest as expected in case of strain partitioningalong a fault located within the arc or the back-arc. A swarm of earthquakes occured close tothe seismogenic interface near Taiwan that might indicate consistent slip vectors deflected byabout 20° from the plate convergence vector. Because these earthquakes are located 40 kmarcward of the mapped transcurrent fault, their deflection can not be simply explained by themotion along this fault. We propose to relate part of these events to the activity of an E-Wtrending tear fault within the subducting slab. Lateral together with vertical motion along thetear can be caused by the nearby arc-continent collision.

West of 122°40'E, the low-taper accretionary wedge is sheared in a direction subparallel tothe convergence vector with respect to the Ryukyu Arc. The bayonet shape of the southernRyukyu Arc slope partly results from the recent (re)opening of the southern Okinawa Trough ata rate of about 2 to 4 cm/yr. Right-lateral shearing of the sedimentary forearc with respect to thenonlinear Ryukyu backstop generates trench-parallel extension in the forearc sedimentsequence at dilational jogs and trench-parallel folding at compressive jogs (Fig.2).

Fig.2: Perspective cartoon illustrating the geometry of releasing andrestraining bends of the toe of the rigid backstop underlying the forearcbasin sediments.


54

STRAIN PARTITIONING BEHIND THE TRENCH TO STRIKE-SLIPTRANSITION: ACTIVE INTRA-ARC DEFORMATION IN OREGON ANDMEXICO

R.M. LANGRIDGE and R.J. WELDON II

Tectonic deformation in the region "behind" the trench to strike-slip transition adjacent to theMiddle America (Mexico) and Cascadia (Oregon) margins is characterised by normal faulting.Seismic moment release is dominated by large (M ~7) earthquakes on faults which scale inlength and displacement to these magnitudes. The latest Quaternary sedimentary records inthese settings have a significant input from their nearby volcanic arcs. The Trans-Mexican Volcanic Belt (TMVB) has experienced at least two of these M7 events inthe last 450 years, including the 1912 Mw 7.0 Acambay Earthquake. We have investigatedthis event in the Acambay Graben (AG) and conclude that: i) this was a typical event for theAcambay-Tixmadeje Fault (ATF); (ii) the recurrence interval for earthquakes on the ATF is~4000 years; (iii) extension across the AG is ~0.25 mm/year and is ~perpendicular to theffabric of the TMVB; and (iv) when scaled to the entire TMVB the historic record for moderateto large events is near complete. The Cascade and Central Oregon (COZ) Zones represent intra-arc and "behind-arc"extensional systems in Oregon. These zones have rates of deformation of 0.1-1.0 mm/yeareach and probably also release seismic strain through large (M ~7) events. Investigation of atephra-rich lacustrine sequence at the Ana River Fault (ARF) in the COZ has revealed alongpaleoseismic record (<80 kyr) with: (i) earthquake recurrence of 10-20,000 years; (ii) a sliprate of 0.1-0.2 mm/year; and (iii) the recognition and understanding of sub-lacustrine eventhorizons for normal faults from near- and far-field localities. Despite low slip rates and long recurrence times compared to subduction zones and strike-slipfaults the "behind-arc" setting should be appreciated as a region of active tectonic deformationwith significant seismic hazard potential. We believe that part of the reason for underestimationof hazard in these zones is related to the lack of microseismic to moderate-sized events, whichmay be a function of a higher heat flow regime in arc-adjacent settings.


55

Fast Subduction, Slow Subduction, and Strike-Slip Faulting, Chile MarginTriple Junction

Stephen D. Lewis

Geology Department, California State University, Fresno sl [email protected]

The Chile Ridge, an active spreading center between the Antarctic and Nazca plates, isbeing subducted beneath South America at 46° S, forming the northward-migrating ChileMargin triple junction. A comprehensive suite of marine geophysical observations, includingscientific drilling during ODP Leg 141, has shown that the forearc of Southern Chile isundergoing subduction erosion where the active Chile Ridge spreading center is presentlybeing subducted. These data also show that Pleistocene-aged sediment is still being accreted atthe base of the trench slope only 30 km north of the site of ridge subduction, indicating that thisregion of the forearc remains unaffected by the triple junction.

South of the triple junction, where the spreading ridge was subducted about 6 my. ago,large folds and landward-dipping thrust faults characterize the structure of the margin. Thisindicates that subduction accretion is now dominating the tectonics of the forearc. If LateMiocene/Pliocene ridge subduction resulted in forearc subduction erosion here, then the well-developed accretionary structures present must have formed in the last 6 my. Thus, thetransition from the dominance of subduction erosion at the triple junction to subductionaccretion following ridge subduction along the Chile Trench occurs over an along-strikedistance of about 50 km, and within time periods of no more than 6 my.

The Linquine-Ofqui fault (LOF) is present landward of the Chile Trench, trending roughlyparallel to the continental margin. Geological and geophysical data suggest that the LOF is aright-lateral strike-slip fault, with deformation beginning in the Oligocene, when theconvergence between the Nazca and South American plates was highly oblique. There isevidence for recent activity on the southern LOF. The LOF appears to be characterized bysplays from the main fault zone that curve seaward into at least two prominent embaymentsunderlain by extensional sedimentary basins in the Chile margin in the vicinity of the triplejunction. The present-day plate convergence direction between the Nazca and South Americanplates north of the triple junction is oblique to the margin, while the convergence directionbetween the South American and Antarctic plates south of the triple junction is nearlyorthogonal to the margin. The northward-migrating triple junction, and hence the northward-migrating transition between oblique and orthogonal subduction, may be related to recentactivity on the LOF and the extensional deformation along its seaward terminations. Thesouthern termination of the LOF is comprised of fault splays that accommodate localizedextensional deformation within the Chile Trench forearc in the Golfo de Penas and otherembayments. Hence, the extensional and subsidence histories of these shelf “forearc” basinsmay provide quantitative constraints on the timing and rates of strike-slip offset along the LOF.As the Chile Triple Junction migrates northward along the South American margin, thesouthern end of active strike-slip faulting that accommodates the oblique component ofsubduction should also terminate.


56

GPS Preliminary results on Northeastern Caribbean plate boundary zone.

A.López (1), P.Jansma (1,2), E.Calais (2), C.Demets (3), P.Mann (4), T. Dixon (5),G.Mattioli (1,2)

1:Department of Geology-University of Puerto Rico-Mayaguez2: CNRS-Geosciences Azur, Sophia Antipolis, France3: University of Wisconsin-Madison4: Institute for Geophysics-University of Texas-Austin5: Rosenttiel School for Marine and Atmospheric Sciences, University of Miami

The North American-Caribbean plate boundary near Hispaniola, Puerto Rico and the VirginIsland is characterized by a complex deformation zone, whose NS width ranges between 200and 250 km. Data from GPS (Global Positioning System) geodetic networks in thenortheastern Caribbean point to differences in strain accommodation in Hispaniola and PuertoRico, which are broadly consistent with geologic and geophysical observations. Explanationsfor the transition in deformation style along the boundary from west to east focus on collisionof the northeastern Caribbean plate boundary with the Bahamas Platform (e.g. McCann andSykes, 1984; Mann et al., 1995). Measurements obtained from campaigns beginning in 1994with the funding of NSF and NASA to create CANAPE (Caribbean-North American PlateExperiment) have been essential for the on-going densification of the Caribbean plate motion. Dixon et al. (1998) presented initial results from CANAPE, which showed a complexlydeforming zone across the Caribbean, as represented by a site at southern Dominican Republic(which is assumed to have velocities of stable Caribbean plate), with 24 + 3.6 mm/yr in aroughly east-west direction relative to stable North America. Although the azimuth is similar,the velocity is twice as fast in comparison with the NUVEL-1 model (Demets et al., 1990).Although the data was processed in ITRF ’94 (International Terrestrial Reference Frame), newmeasurements from 1998 that were obtained in late 1998 were processed with two differentsoftwares and with ITRF’96, which yielded results with more accuracy in a span of eight years(1990-1998). These updated velocities of all sites in the Dominican Republic still decreasesfrom south to north, but the azimuths now consistently show a component of convergenceacross the major strike-slip both onshore an offshore. With respect to Puerto Rico, a site in thenorth has similar motion to the southern station in the Dominican Republic at a little more than16 mm/yr with respect to fixed Turk island of the Bahamas, and to fixed North American plate.However, these results does have a statistically significant northward velocity of a few mm/yrwith respect to the southern Dominican Republic, implying increased convergence along theboundary north of Puerto Rico and a component of left-lateral displacement between northernPuerto Rico site and Dominican Republic southermost station.

References:

Demets, C., R.G. Gordon, D.F. Argus, and S. Stein, Current plate motions, Geophys J. Int., 101, 425-478,1990.

Mann, P., F.W. Taylor, R. Edwards, and T.L. Du, Actively evolving microplate formation by oblique collisionand sideways motion along strike-slip faults: An example from the Northeastern Caribbean plate margin,Tectonophysics, 246, 1-69, 1995.

McCann, W.R., and L.R. Sykes, Subduction of aseismic ridges beneath the Caribbean plate: Implications forthe tectonic and seismic potential of the northeastern Caribbean, J. Geophys. Res., 89, 4493-4519, 1984.


Part 4: Gravity data and seismic tomography from the southeastern Carpathians:Imaging the terminal phase of subduction

Frank P. Lorenz1, Michael Martin1, Blanka Sperner1, Friedemann Wenzel1, VictorMocanu2

1 Geophysical Institute, University of Karlsruhe, Hertzstr. 16, 76187 Karlsruhe, Germany2 Geophysical Institute, University of Bucharest, 6 Traian Vuia Street, 70139 Bucharest 1, Romania

GravityA zone of negative Bouguer anomaly parallels the southeastern Carpathian arc (Fig. 6).

Lowest values (-110 mgal) are recorded from the Getic depression (for loc. s. Fig. 4) whereoverthrusting of the Tisia-Dacia block onto the northern margin of the Moesian platform causedformation of deep sedimentary basins. In contrast to this upper-crustal feature, an offset in theBouguer anomaly just north of the SE-bend of the Carpathian arc seems to be caused by „deeperprocesses“: the offset separates the region of thickened crust in the N from the region of basinformation and fast foredeep subsidence in the S (Fig. 4 & 6). Additionally, it represents thenorthern border of the high-velocity body outlined by seismic tomography.

Fig. 6: Bouguer anomaly of Romania / southeastern Carpathians

44

46

48

28262422

mga l

-140-120

-100-80-60-40-20

0

2040

Seismic tomographyA 3-D model of the velocity-depth structure for the SE-bend of the Carpathian arc was

calculated by inverting teleseismic P-wave travel-time residuals. The digitally recordedwaveforms for 125 events were picked, travel-time residuals were calculated with the IASP91reference tables and the inversion was done with the ACH-method. The weighted relativeresidual, which are more or less independent of hypocenter localisation errors, were minimizedusing a standard least squares algorithm. Our current best fit model with a variance improvementof 70% reveals a local high-velocity body with maximum perturbations of about 3% which istotally embedded in a low velocity medium with negative peak perturbations of about 5%. Allintermediate depth earthquakes lie within the high-velocity body. The seismically quiet depthrange (40 to 70 km) is (within the limits of resolution) not associated with a velocity anomaly. Ona global scale, all images of active subduction zones are characterized by high P-wave velocities.We interpret this as strong evidence for remnants of a slab of oceanic lithosphere beneath thesoutheastern Carpathians.

The high-velocities body displays a peculiar structure, the main feature of which is a change ofstrike from top to bottom. The uppermost high velocity material strikes roughly SW-NE (Fig. 7).This is consistent with subduction to the NW or delamination and rollback to the SE. Deeperportions of the high velocity material, however, are oriented more N-S, close to the strike of theEastern Carpathians. Presumably this was also the strike orientation of subduction before the lastslab segment was totally incorporated into the mantle (Fig. 2c). Thus the directional change ofsubduction during the last phases of collision in the Eastern Carpathians may be preserved in thetomographic image. N-S oriented high-velocity material would represent older westwardsubducted material that is detached from the foreland lithosphere, but still attached to the upperSW-NE trending part of the slab. This deeper (N-S striking) portion of dense oceanic lithospheregenerates a downward force which increases shear stresses in this passive slab.

Fig. 7: Seismic tomography of the upper mantle between 112 and 152 km in depth. For 1999 a large tomo-graphic experiment with 143 stations and a reasonable resolution towards depths of 400 km is planned(CALIXTO = Carpathian Arc Lithosphere Cross-Tomography Project).


59

Lithospheric-Scale Modeling of Transpressional Regimes: Application to theSouth Island, New Zealand

R.Malservisi and K.P.Furlong

Geodynamics Research Group, Department of Geosciences, Penn State University, University Park, PA, USA.e-mail [email protected]

The pole of rotation of the Indo-Australian and Pacific Plates is close to the plate boundary inthe area of New Zealand; this makes the dynamics of the region very sensitive to any change inplate motion. A migration of the pole of rotation westward characterizes the history of the Indo-Australian/Pacific plate in the last 6(10?) Ma. This leads to an increasing compressionalcomponent in the relative motion along the Alpine and Puysegur faults and a change from atranslational to a transpressional regime. The effects of this change are very pronounced in the Fiordland region where the lithosphericinteractions have led to a subduction like regime manifested by deep earthquakes (deeper than150 km), large gravitational anomalies and rapid uplift. The response of the lithosphere to this evolution in plate kinematics is the principal feature weare testing in our 3d-FEM model (TECTON). In particular we have tested the conditions thatinfluenced the early stage of the transition to generate the present situation of a localized regionof deep earthquakes and significant bouguer gravity anomaly. The model is focused on the area of Fiordland and it is parametrized to evaluate the responseof the lithosphere to different geometries, rheologies and thermal states, to verify how theseparameters can affect geophysical observable such gravity, topography and seismicityproviding an important constraint on dynamics of the area.


60

Classification and tectonic comparison of subduction to strike-slip transitionson active plate boundaries

Paul Mann and Cliff Frohlich

Institute for Geophysics, University of Texas at Austin, 4412 Spicewood Springs Road, Bldg. 600, Austin, TX78759

Strike-slip to subduction transitions on active boundaries occur where Benioff zones abruptlyor gradually terminate on crustal strike-slip zones. In ancient rocks, these transitions arerecorded in island arc rocks that are superimposed by strike-slip faults and related rocks withno evidence for continued arc activity. In this meeting, we will discuss the active and ancientexamples together with the aim of identifying the main tectonic, seismogenic, and geologicelements of subduction to strike-slip transitions. To initiate this five day discussion, we classify transitions on the world's active plateboundaries into three types: Open corner type: This type of transitional boundary is characterized by plate boundariesdescribing obtuse "open" angles when viewed from the downgoing plate (Figure A).Characteristics of open corner transitions include: 1) the trench and Benioff zone vary smoothlyalong strike; 2) the trench is anomalously deep with depths of 6-9 km; 3) the Benioff zonesteepens to vertical as the strike-slip boundary is approached; 4) discontinuous pieces ofsubducted slabs are present in the area of the oversteepened Benioff zone; 5) strike-slip faultsare present above the area of detached slabs and lengthen in a trenchward direction; and 6) acommon tectonic setting of open corner transitions is a closing "loop" or strongly arcuate islandarc constrained between two continents; inactive arc segments mark already collided areas nowconverted to strike-slip faults and active arc segments mark actively subducting areas (FigureA). The tectonic process responsible for open corner transitions appears linked to collision ofan unsubductable bathymetric high at an adjacent closed corner (continental parts of North andSouth America plates in Figure A). Collision of the high and the arc detaches the alreadysubducted oceanic slab from the high and allows the slab to sink into the mantle at a steep tovertical angle. Strike-slip faults mark the suture between the unsubductable high and thecollided arc and lengthen trenchward with continued progressive collision (Figure A). Discussed examples of open corner transitions: Northeastern and southeasternCaribbean (Figure A), northeastern and southeastern Scotia, central Aleutians, northern andsouthern Marianas, Philippines, Taiwan, San Cristobal trench of Solomon Islands, southernVanuatu, northern Tonga, North and South Islands of New Zealand, southern CarpathianMountains, Aegean, and Sumatra-Andaman. Closed corner type: This type of transitional boundary is characterized by plateboundaries describing acute "closed" angles when viewed from the downgoing plate (FigureB). Characteristics of closed corner transitions include: 1) a narrow seafloor bathymetric highor continental block adjacent to the closed corner on the downgoing plate (e.g., Hawaii-Emperor seamount chain-Obrutchev Rise in Kamchatka - Figure B); 2) the trench and Benioffzone conforms to the shape of the incoming colliding feature; 3) the Benioff zone is locallycontorted into a sharp angle but is not necessarily vertical in dip or associated with detachedpieces of slab; and 4) strike-slip faults may not be present on both sides of the closed corner.The tectonic process responsible for the transition appears linked to the subduction of thenarrow, bathymetric high that continues to subduct despite the formation of the prominentcorner or cusp in the subduction geometry (Figure B). Because subduction of the highcontinues, detachment and sinking of an oceanic slab from the high may not occur. Discussed examples of closed corner transitions and associated collidingfeatures: Kamchatka (Hawaii-Emperor seamount chain - Figure B), western Himalayas(Indian sub-continent), southeastern Alaska (Yakutat block)


61

Possible relation between open and closed corner transitions. Collision of thebathymetric high and the arc forms a prominent cusp in the subduction zone (Figure B).Continued interaction of the high, the arc, and lengthening strike-slip zones could produce anopen corner adjacent to a closed corner (e. g., open corner transition of central Aleutians isadjacent to closed corner transition of Kamchatka). Therefore, the collision process at closedcorner transitions may initiate the eventual formation of adjacent open corner transitions. Discussed examples of adjacent open and closed corner transitions. Kamchatkaand central Aleutians; northern Marianas and Ogasawara Plateau; southern Marianas andCaroline Ridge; western Himalayas and western Indian sub-continent; northeastern Scotia andNortheast Georgia Rise; northeastern Caribbean and Bahama-Greater Antilles collision zone;southeastern Caribbean and northern South America collision zone. Triple junction type. This is the most abrupt type of strike-slip to subduction transitionand consists of a fault-fault-trench (FFT) or fault-trench-trench (FTT) triple junctions asdescribed by McKenzie and Morgan (1969) (Figure C). FFT junctions are stable if the trenchand one fault are colinear. As plate motions proceed, the junction maintains its geometry andboundary types. When the trench and one fault are not colinear, the junction is unstable and atriangular gap appears at the junction that must be accommodated by surrounding regions(Figure C). Ridge-trench-trench (RTT) junctions appear shortlived in most cases becauseridges often cease spreading on entering the trench and convert to strike-slip faults and FTTjunctions. Discussed examples of triple junction transitions: Gautemala (FTT); Panama(FTT); Chile (RTT where ridge appears to continue spreading on subduction), Mendocino(FTT - Figure C), Woodlark (RTT to FTT because ridge appears to have converted to a strike-slip zone on subduction); New Ireland (TTF): New Guinea (FTT).


63

The role of Paleogene volcanism in the evolution of the northern Caribbeanmargin

Giuseppina Mattietti-Kysar

The George Washington University, 2029 G St. NW, Washington D.C. 20052, USA

Since Cretaceous time, the northern Caribbean margin has evolved from subduction controlledplate boundary to the present day strike-slip tectonic regime. This process has probably involveddifferent, contemporaneous tectonic settings along the strike. This study focuses on the Paleogene age magmatism of the Sierra Maestra (southeastern Cuba),a 250 km long, east-west striking structure, that represent the axial part of a former volcanic arcwith northward subduction (Lewis and Draper, 1986). The Sierra Maestra is in alignment with thesubmerged Cayman Ridge to the west, and it is bounded to the south by the Oriente deep and onland by two tectonic depressions (Cauto valley and Guantanamo basin). The Sierra Maestra exposes a thick, relatively undeformed succession of volcanic andvolcaniclastic rocks, intercalated with marine sediments. Several sets of dikes cut trough thissuccession, and small plutonic bodies crop out along the southern flank of the Sierra and on thehighest peaks. The Paleogene volcanic rocks rest with discordance on reworked, moderatelymetamorphosed material most probably from the weathering of the Cretaceous volcanoes to thenorth (Iturralde-Vinent, 1996). This study supports the idea that the Sierra Maestra was part ofthe Cayman arc structure, as suggested by Perfit and Heezen (1978) who pointed out thesimilarities between lithologies observed in the Sierra Maestra and those dredged on theCayman Ridge. During three field seasons (1996-1998) sampling of magmatic rocks was carried out forpetrological and geochemical studies. New analytical results show that the Paleogene arcvolcanism followed a tholeiitic differentiation trend. Discrimination diagrams for the volcanicrocks based on Zr, Ti, Y, Ta, Th, Yb, etc ... indicate a primitive, oceanic arc signature. REE plotsreveal depletion of the magmatic source with respect to N-MORB. In addition the data for theintrusive rocks show the existence of two distinct groups with depleted patterns. These twogroups correspond with the western and the eastern plutonic bodies respectively (Kysar et al.). U-Pb zircon ages determined on seven samples from the Sierra Maestra span a 10 Ma interval(from 56 to 46 Ma). The oldest intrusion, located toward the western end of the Sierra Maestra, isseparated by a 6 Ma hiatus from the younger intrusions to the east. This suggests the existence oftwo magmatic episodes in the studied area. These absolute age determinations confirm theexistence of a 20 Ma hiatus from the last dated Cretaceous volcanism in Cuba at the end of theCampanian (Kysar et al., 1998b). None of the analyzed fractions show any inheritance fromolder zircons, thus confirming the juvenile character of the Sierra Maestra arc activity. Theprimitive, depleted geochemistry and the absolute age dating indicate that the Sierra Maestramagmatism was not simply the final stages of the Cretaceous arc activity, although it wasprobably built on the remains of the Cretaceous forearc (the metasediments mentioned above). This study also shows that the Sierra Maestra arc was short-lived and with at least two magmaticepisodes, each from a chemically different source. Although it is not yet possible to constrain theolder magmatism (56 Ma) to a specific tectonic episode, it is concluded that since 56 Ma (LatePaleocene) the western part of the active margin of the northern Caribbean was situated along theeast-west striking Cayman-Sierra Maestra arc. This arc was striking at about 45 degrees from thenorthwest-southeast Cretaceous arc. Probably such counterclockwise rotation began as soon asthe Yucatan continental block collided against the western edge of the Cretaceous arc. This eventmust have terminated the Cretaceous volcanism in Cuba, opened the Yucatan back-arc basin(Wadge et al., 1984) and triggered the onset of Cayman-Sierra Maestra arc.


64

The younger magmatism ( Eocene, 50-46 Ma) was active until the culmination of collision withthe Bahamas continental platform. This second collision event most probably changed the stressconfiguration of this area, transferring to tangential stress in order to accomodate the Caribbeancrust which was now moving eastward. This event also resulted in the emplacement of theMayari-Baracoa ophiolite complex in the mid-late Eocene (as confirmed by the age of theolistostromes to the north of the Sierra). As a consequence, uplift of the Sierra Maestra (asubmarine arc) occurred, and the Cauto and Guantanamo basins were formed. The present datagive no clear information on the time of the opening of the Cayman Trough.

AcknowledgementsThis presentation has been made possible thanks to a grant-in-aid for junior scholars

offered by the Cosmos Club Foundation of Washington D.C.

ReferencesIturralde-Vinent, 1996. Cuba: el arcipelago volcanico Paleoceno-Eoceno medio. In: Cuban ophiolites and volcanic arcs: First Contribution IGCP Project 364: 231-245 p.Kysar, Green, Perez, Mendez, and Rodriguez, 1998a. New Geochemical data for the igneous rocks of the Sierra Maestra, Southern Cuba. Abstract, 15th Caribbean Geological Conference, Contributions to Geology, University of West Indies , #3.Kysar, Mortensen, Lewis, 1998. U-Pb zircon age constraints for Paleogene igneous rocks of the Sierra Maestra, southeastern Cuba: implications for short-lived arc magmatism along the northern Caribbean margin. Abstract, GSA Meeting, 1998.Lewis and Draper, 1990. Geology and tectonic evolution of the northern Caribbean margin. In: The Geology of North America, vol. H, The Caribbean Region, p. 77-140.Perfit and Heezen, 1978. The geology and evolution of the Cayman trench. GSA Bulletin, v. 89, p. 1155-1174.Wadge, Draper and Lewis, 1984. Ophiolites of the Northern Caribbean, a reappraisal of their roles in the evolution of the northern Caribbean Margin. In Ophiolites and oceanic lithosphere, p. 367-380.


65

Strain partitioning and development of upper plate strike-slip faults duringoblique convergence: An example from Sumatra

Robert McCaffrey

Department of Earth and Environmental Sciences, Rensselaer Polytechnic Institute, Troy NY 12180-3590,Phone: (518) 276-8521 Fax: 276-8627, Email: [email protected], Web: www.rpi.edu/~mccafr

Margin-parallel strike-slip faults are frequently found inboard of oblique convergentmargins and there is little doubt that a direct mechanical relationship exists between the shearfaulting and the shear stress along the dipping thrust fault. We seek to understand the details ofthe mechanics because the resulting geology, such as the location of the strike-slip fault, willreveal in some fashion the forces involved. In this talk we will focus on the question of whereand how strike-slip faults develop in the overriding plate and what this tells us about theinterplate forces.

At the subduction zone of northern Sumatra, GPS measurements reveal that the strainassociated with the oblique plate motion is fully partitioned between trench-normal convergencewithin the forearc and arc-parallel shear strain within 30 km of the Sumatra fault. The strainpartitioning is consistent with kinematic inferences made from earthquake slip vectors althoughthe slip vectors indicate slightly less partitioning. GPS measurements at several other plateboundaries are used to estimate the degree to which strain is partitioned during obliqueconvergence. Principal strains within the leading edge of the upper plate are compared to theazimuths of earthquake slip vectors, plate convergence vectors, and the orientations of thedeformation fronts. At some margins, such the Himalayas (Nepal), Nankai, and Sumatra, thecontractional principal strain direction is perpendicular to the deformation front even thoughplate convergence is very oblique. Margin-parallel shear strain is then localized landward of thestrongest contractional strain. In other places, such as the large bend in the Peru-Chile trenchfrom 15°S to 18°S, where the obliquity is 35°, in Alaska (Prince William Sound) whereobliquity is 20°, and in Ecuador where obliquity is 25°, the contractional strain aligns with theplate convergence vector.

Finite element modeling is being used to try to understand the conditions under whichstrain is partitioned or not. In particular, we are addressing the factors that can cause theprincipal strain directions rotate by up to 45° between the forearc and the arc regions ofconvergent margins. In some cases, such as Sumatra, the presence of an arc-parallel strike-slipfault may promote the strain partitioning but in general, we find that a strike-slip fault need notexist for partitioning to occur. Moreover, the models suggest that the strike-slip faults developbecause of strain partitioning within the upper plate but they need not develop at the volcanicarc.


66

Space and Time Variance Among Large Earthquake Ruptures Inferred From aWorldwide Data Set

JAMES P. McCALPIN

GEO-HAZ Consulting, Inc., P.O. Box 1377, Estes Park, Colorado, 80517, USA

Time Variance In 1997-98 McCalpin, Slemmons, and Nishenko inventoried the published and "gray"literature for paleoearthquake chronologies in which 3 or more consecutive paleoearthquakeswere dated by numerical means. We found 41 chronologies containing a sum of 161paleoearthquakes; most of these local sequences contained 3 or 4 dated paleo-earthquakes, i.e.2 to 3 dated recurrence intervals (RIs). The variability in recurrence was assessed bynormalizing the individual recurrence times (T) in a local sequence to the average recurrence(Tavg) of that sequence. The frequency distributions all have a mean=1 (since the data were allnormalized), and thus the sigma value is also the Coefficient of Variation (COV).

Table 1. COV of T/Tavg for Various Types of FaultsFault Type No. of RIs COV of T/TavgStrike-slip 44 0.39Subduction 45 0.39Normal 56 0.32Reverse 16 0.37ALL TYPES 161 0.36

All Fault Types Grouping all the 161 normalized RIs together, the frequency distribution is slightlyasymmetric, with a mode of 1.0- 1.1, and more RIs smaller than Tavg than larger. Theextremes are T/Tavg= 0.2 and 2.0. Thus, if a fault had a long term mean RI of 1000 years, andit obeyed the grouped frequency distribution, its RIs over time would range from 200 years to2000 years, with RIs shorter than the mode being slightly more common than the largerintervals. This latter tendency implies weak clustering of earthquakes in time. Clustering resultsin more short (intra-cluster) RIs and fewer long (inter-cluster) RIsReduction of Variance with More Paleoearthquakes The 13 local sequences containing only three dated paleoearthquakes (two RIs) had COVsranging from 0.02 to 0.96, but as the number of RIs in the sequences increases, the spread ofCOV values decreases, converging on a value of ca. 0.3. One interpretation of this pattern isthat all the sampled faults have a long term COV of recurrence of ca. 0.3, and that the larger aportion of their complete recurrence history has been sampled by the paleoseismicinvestigation, the closer the estimated COV is to the real COV. If seismogenic faults all share afundamental property of generating earthquake time-series with a recurrence COV=0.3-0.4,this fact would have critical implications for seismic hazard forecasting. Spatial Variance

We also inventoried 56 historic surface ruptures where post-faulting investigations hadyielded at least 15 high-quality measurements of one or more components of slip. Ruptureswith 15-30 slip measurements were most common, with a decreasing frequency for moremeasurements. The largest number of slip measurements (269) was made on the 1920 Haiyun,China earthquake. One way to characterize the frequency distribution of slip along strike is to compare theaverage displacement (AD) to the maximum displacement (MD). This ratio ranges from 0.30 to0.42 for various fault types, with an overall average of about 0.35. Thus, the average


67

displacement tends to be about 35% of the maximum displacement, or put another way, themaximum displacement tends to be about 2.9 times larger than the average displacement. Thisfinding contrasts with the conclusion of Wells and Coppersmith (1994), who concluded thatAD averaged about 50% of MD. Our results indicate that the maximum displacement on arupture is even more anomalous than previously suspected.

Table 1. Ratio of Length-Weighted Average Surface Displacement (AD) to Maximum SurfaceDisplacement (MD), for Various Fault Types

AD/MDComponent of Displacement

Fault Type Dvert (N) Dss (N) Dnet (N) Strike-slip - 0.38±0.08 (18) -Normal 0.33±0.09 (11) - -Normal-oblique 0.31±0.07 (6) 0.37±0.11 (2) 0.42±0.11 (2)Reverse 0.38±0.09 (6) - -Reverse-oblique 0.33±0.12 (16) 0.30±0.11 (14) 0.35±0.10 (11)

Based on measurements of displacement in historic surface ruptures, it appears that normalfaults have the largest proportion of small displacements relative to the maximum (i.e., <10%Dmax). Strike-slip ruptures, in contrast, have the highest proportion of larger displacementsrelative to the maximum. Reverse fault ruptures show intermediate tendencies between thesetwo extremes. The high proportion of small displacements in normal faulting events means thatsuch ruptures are the most affected by erosional modification obscuring smaller displacements.This may explain why paleoseismic evidence for normal faults often indicates anomalouslyshort rupture lengths for the amount of displacement measured. The most reasonableexplanation is that post-faulting weathering has obscured a significant part of the originalrupture length. This phenomenon would not affect reverse or strike-slip ruptures as strongly. Using the frequency distributions, one can make a statistical prediction of Dmax given one ormore random measurements of displacement during a paleoearthquake. If the measurement sitewas not randomly located with respect to visible displacements from the studied event, it ispossible to use only part of the frequency curve. Using such statistical methods , it is no longernecessary to assume that displacements measured in trenches were the largest (Dmax) for eachrecognized paleoearthquake.


68

Global overview of earthquake hazards at transition zones

James P. McCalpin

GEO-HAZ Consulting, Inc., P.O. Box 1377, Estes Park, CO 80517 USA

Historic earthquakes in transition zones have caused damage from both direct, on-land strike-slip style tectonics and indirect, offshore subduction-style tectonics. In a way, transition zonesare vulnerable to a wider suite of earthquake hazards than transcurrent or subduction plateboundaries alone.Surface Rupture: Some of the largest historic surface displacements on strike-slip faultshave occurred in transition zones. For example, the 1855 West Wairarapa, New Zealandearthquake may have been accompanied by as much as 12-15 m of dextral offset, or twice themaximum offset in the 1906 San Francisco, California earthquake. However, even decimetersurface displacements have caused collapse of buildings. In the 1972 Managua, Nicaraguaearthquake, the 3-story reinforced concrete Customs House was totally collapsed by sinistraldisplacement of 28 cm. Sub-decimeter offsets caused widespread collapse of woodframe/adobe houses along the fault traces.

The most destructive effects of surface rupture are often on lifelines and utilities.Following the 1972 Nicaragua earthquake fires raged for weeks, because water mains had beenbroken by surface faulting.Ground Shaking/ Amplification: Ground shaking from on-land strike-slipearthquakes tends to have more energy concentrated in the high-frequency bands than shakingfrom more distant, offshore subduction earthquakes. Because transition zones experience bothtypes of events, a wider spectrum of building types may have resonant frequencies that matchthose of the seismic wave train. Thus, high-rise buildings in transition zones may experiencesevere damage due to long-period waves coming from great distances (e.g., 1985 Mexico Cityearthquake), as well as due to short-period waves from smaller but closer strike-slipearthquakes.Liquefaction: The strength of ground shaking in transition zones is certainly sufficientto cause liquefaction, if the local materials are susceptible. Liquefaction in the 1692 Jamaicaearthquake destroyed the city of Port Royal and led to the founding of modern Kingston.Liquefaction and building settlement during the 1960 Niigata, Japan earthquake is an often-cited classic case of bearing load failures. In the Caribbean, liquefaction occurred during the1976 Guatemala earthquake and the 1997 Venezuela earthquake. Both prehistoric and historicliquefaction has been documented for the Septentrional fault, Dominican Republic, by recentstudies of Tuttle, Prentice, and Pena. However, the 1972 Nicaragua earthquake was notaccompanied by significant liquefaction, due to water table depths of 10-30 m in Managua andthe coarse, angular nature of pyroclastic deposits there. As in most areas, liquefactionsusceptibility is controlled by local sedimentary environments, and microzonation maps canonly be produced from detailed Quaternary geologic mapping combined with analysis ofborehole data.Landslides: The topography in many transition zones is rugged, and near-angle-of-reposeslopes are susceptible to earthquake-induced landsliding. For example, the 1958 Fairweather,Alaska earthquake (Mw=7.7) caused landslides over an area of ca. 70,000 sq. km. The Mw7.5 Guatemala earthquake of 1976 caused slides over about 18,000 sq. km. These landslidestypically obstruct roads and railroads, hampering relief efforts, and sometimes contributing thelargest share of monetary earthquake damage. Although the low-relief terrain of cities intransition zones are not hit particularly hard by landslides, substandard housing on hillssurrounding metropolitan areas is becoming increasingly vulnerable to landslide damage.


69

Lateral-spread landslides form a special case associated with subsurface liquefaction,and have occurred on gently-sloping terrain in metro areas. Much of the damage to the docks atKobe, Japan in 1995 was caused by lateral spreading of man-made land in the harbor. Thistype of damage may be expected in any urban harbor area where bays were filled withhydraulic slurry material or nonstructural fill placed at low densities.Geodetic Changes: Subduction earthquakes are typically accompanied by permanentchanges in coastal elevations, with landward belts of subsidence and seaward belts of upliftparalleling the subduction zone. In the 1931 Napier, New Zealand earthquake maximum upliftalong the coast was 2.5 m; the 1994 Petrolia, California earthquake experienced coastal upliftup to 1.4 m. In areas of low tidal ranges, meter-scale uplifts can render harbor facilitiesunusable.Tsumanis: Two catastrophic tsunamis occurred in 1992 in or near transition zones, andhighlight this hazard. The 1 Sept. 1992 Nicaragua earthquake (M 7.0) generated waves 8-15m high that struck 26 towns along 250 km of Nicaragua's Pacific coast. The waves penetratedas much as 1 km inland, killing 116, with 63 missing, 489 injured, and 40,000 homeless.Total losses were US$ 25 million. The 12 Dec. 1992 Indonesia earthquake (Ms 7.8) createdwaves with a maximum runup height of 26.2 m with a landward penetration up to 300 m,which severely scoured some coastal areas and deposited up to 1 m of sediment elsewhere.About 1000 people were killed, 500 injured, and 90,000 left homeless.Slip Partitioning and the Earthquake Cycle

For hazard forecasting it would be helpful to know whether subduction and strike-slipearthquakes alternate in a transition zone, or whether sequences of gap-filling subductionevents follow sequences of gap-filling strike-slip events. The historic record in Hispaniolaincludes two consecutive strike-slip (?) events in 1842 and 1887 in western Hispaniola,followed by two consecutive subduction events in 1946 and 1953 in eastern Hispaniola. Thus,at present there are two gaps, one on the eastern Septentrional fault, another on the westernNorth Hispaniola Deformed Belt. Which of these zones will fail next? This question could beaddressed by boundary-element modeling, which predicts the increase of static stresses (andthus, time-to-failure) on fault-bounded crustal blocks caused by slip on adjacent fault blocks.


70

Collisional Tectonics of the Bahama Bank along the Northern Margin of theCaribbean Plate: Cause and Effects

William R. McCann

Earth Scientific Consultants, 6860 West 99th Avenue, Westminster, Colorado 80021

Several Late-Miocene to Recent tectonic events along the northern margin of theCaribbean - North American Plate Boundary are related to the entrance of the Bahama Bankinto Northern Lesser Antilles or Puerto Rico subduction zone. The reaction to the collision ofthis buoyant feature with the subduction zone depends upon the obliqueness of subduction,and the buoyancy of the interacting Bank.

In the Northern Lesser Antilles, the Bank enters the subduction zone about MiddleMiocene, coinciding with a period of magmatic quiescence along the northern part of thesubduction zone. Magmatism appears again in earliest Pliocene time, but with its locus shiftedwestward away from the Lesser Antilles Trench. This shift occurs because of erosion of theforearc of the Northern Lesser Antilles, or by a shallowing in the dip of the subducted platecaused by a more buoyant slab or both. The event affects about 300 km of arc.

Along the more oblique margin to the North and West, the effects occur later and are ofa different form. Puerto Rico suffered a major tectonic event in Late Miocene- Pliocene Time.The Bahama Bank lay just north of Puerto Rico during this time. Shallow water limestones ofPliocene and older age are now found at 4 km depth along the northern flank of the Puerto RicoPlatform. The rapid subsidence of this forearc was probably due to accelerated tectonicerosion.

Also, Puerto Rico experienced a period of accelerated rotation when it rotatedcounterclockwise 24o starting about 11 MA and ceasing about 4-5 Ma. This event wascontemporaneous with: 1- Development a rotating microplate along the Caribbean-NorthAmerican margin, containing all of the Puerto Rico Virgin Islands Platform and perhaps asignificant part of Hispaniola; 2- Rotation induced extension (N30W) along the SE edge of themicroplate i.e. Anegada Passage; 3- Rotation induced compression (subduction) along the SWedge of the microplate i.e. Muertos Trough; 4- Development of a separate microplate westwardof the immediate collision zone (Gonave Microplate), including the initiation of the Jamaicanzone of deformation. This event affects at least 1500 km along the northern margin of theCaribbean Plate and is best characterized as collision induced rotational back-arc thrusting,shifting a segment of the Greater Antilles island arc onto the crust that lay in its back arc region.

In the last 5 Ma, 1- Rotation of the Puerto Rico Microplate ceases and correspondingextension and compression induced by that rotation also ceases around Puerto Rico; 2- A newregion of extension (N45E) develops between the Puerto Rico and Hispaniola. This new areaof extension reflects continuing subduction along the Western Muertos Trough induced bycontinuing collision of the Bahama Bank with Hispaniola, associated lateral arc extensioncaused by the geometry of subduction zones of opposing dips, and docking of Puerto Ricowith the buoyant Caribbean Plate.

If Hispaniola in fact rotated in tandem with Puerto Rico during the 11-4 Ma periodmentioned above, then a significant segment of "Caribbean Plate" that lay to the SW of themicroplate would have been subducted along the Muertos Trough and its extension to the NW.A pole of rotation for the opening of the Anegada Passage was probably located about 100kmsouth off the western coast of Puerto Rico (17N, 67W). If that were in fact a fixed pole forrotation of the Puerto Rico-Hispaniola Microplate, then convergence of 200-300km would berequired along the westernmost portion of the microplate in Haiti, and 100-200km in theDominican Republic. In the Eastern Dominican Republic there is a 100-200km downdip


71

section of seismic activity generally believed to be North American Plate that preceeded theBahama Bank into the Puerto Rico Trench.

We suggest that another possibility exists for the origin of that cluster of intermediatedepth seismicity. Perhaps that earthquake activity represents lithosphere of the Caribbean Platesubducted at the Muertos Trough by the collision induced rotation of the extended Puerto RicoMicroplate. This concept is consistent with deformation observed in Central Hispaniola and theextensive accretionary prism just south of the Dominican Republic. It may also explain theenigmatic gap in seismcity between shallow events, clearly related to present day subduction atthe westernmost portion of the Puerto Rico Trench, and the intermediate depth events beneathEastern Hispaniola. We suggest that perhaps the latter events should instead be related toshallower seismicity near the Muertos Trough. This concept would also predict the existence ofan easterly narrowing strip of the Caribbean Plate lies underthrust beneath the Mona Passageand westernmost Puerto Rico.

If Hispaniola did not rotate along with the Puerto Rico-Virgin Islands Platform, thenthe seismicity gap mentioned above may represent a real break in the subducted NorthAmerican lithosphere. However, this model suggests that 10's if not 100 kilometers ofdeformation occurred somewhere between Western Puerto Rico and Eastern Hispaniola as thePuerto Rico Microplate rotated. No such deformation zone has yet been identified.


72

Crustal Evolution of the Cascadia Subduction Marginin the Vicinity of the Mendocino Triple Junction

Patricia A McCrory1, Douglas S Wilson2, and Mark H Murray3

1US Geological Survey, Menlo Park, CA 94025, 2Geological Sciences Department, University of California,Santa Barbara, CA 93106, and 3Geophysics Department, Stanford University, Stanford, CA 94305

The Mendocino triple junction occupies a transitional zone between the Cascadia subductionregime to the north and the San Andreas transform regime to the south. The continental marginat the southern end of the Cascadia subduction zone is composed of a 15-km-wide deformationfront backed by the Franciscan complex—a 100-km-wide belt of Cenozoic and Mesozoicsubduction-complex terranes juxtaposed against the Mesozoic and Paleozoic Klamath Mountainterrane on the east. The strata that fill the Humboldt basin, a 40-km-wide and 200-km-longsynclinorium overlying the subduction-complex basement, represent a record of the marginbehavior. This record of subduction-related tectonism—when compared with the style ofmodern tectonism—yields insight into the ongoing transition from a subduction regime to atransform regime in the vicinity of the northward-migrating triple junction. Humboldt basin consists of a series of coalesced sub-basins. The southernmost sub-basin isthe Eel River basin which contains a marine stratigraphic sequence that documents a restricteddeep-water depocenter in the Miocene, shoaling as the depocenter filled with turbidite facies inthe Pliocene, and a widespread slope and shelf environment that developed as adjacent sub-basins coalesced in the Pleistocene. This sequence is consistent with continuum modelsexplaining the evolution of accretionary margin basins, yet the vertical tectonic signal preservedwithin this sequence reveals several significant deviations (McCrory, 1989; 1995): (1)Humboldt basin strata overlie Paleogene and older Franciscan Complex rocks, implying ahiatus of about 15 My between accretion of basement rocks and deposition of overlap strata;(2) deep-water conditions during deposition of overlap strata persisted for at least 10 My (ca.14-4 Ma); and (3) rapid uplift began along what is now the southern basin margin late in basinhistory (ca. 3.5 Ma) followed by onset of folding (ca. 2.0-1.5 Ma). This ongoing phase ofuplift and shortening may be driven by the increased coupling between the subducted oceanicslab and overlying subduction-complex rocks, inferred to result from the strongly obliqueconvergence between the Juan de Fuca plate and northern California that began at about thesame time (Wilson, 1993). Thrust faulting began ca. 1 Ma. North of 40.8°N, thrust faults trend north-northwestconsistent with subduction-related shortening within the upper plate. This structural pattern isdisrupted in the Cape Mendocino region where active thrust faults developed in the Yager andCoastal Franciscan terranes (and their offshore equivalents) trend west-northwest. Acomprehensive survey of active thrust faults within onshore Franciscan terranes (McCrory,1996) indicates permanent contraction of about 10 mm/y directed 55±10°. The contractionassociated with these young faults is oblique to the relative motion between the southern Juande Fuca plate and northern California, suggesting that the strain may be distorted by a Pacificplate buttress propagating northward with the migrating triple junction. However, fewwidespread age markers occur in the onshore Quaternary strata, hampering efforts to determinewhether the onset of crustal faulting in the Humboldt region was synchronous, diachronous, orprogressed from south to north as might be expected from the northward migration of theMendocino triple junction. Modern strain accumulation across the Humboldt region (1981-1989) indicates interseismiccontraction of about 15 mm/y directed 35±8° (Murray and Lisowski, 1995). Motion betweenthe southern Juan de Fuca plate and the northern California margin is about 40-50 mm/ydirected 47±7° (McCrory, Wilson, and Murray, 1995). The discrepancy of 12±11° between


73

the direction of interseismic contraction and the direction of subduction-related convergence,although not very statistically significant, suggests that a component of modern deformationmay be unrelated to strain accumulation on the megathrust. This component may be driven byconvergence between the easternmost Pacific plate and the Eel River basin area at about 38mm/y directed 340° (McCrory, Wilson, and Murray, 1995). The onshore Pacific plate buttressmay have formed recently by an inland jump in the local plate boundary, or alternatively, thebuttress may have translated northward along the margin for some time. In either case, theskewed contraction direction likely results from both localized anomalies in Juan de Fucamotion adjacent to the Pacific plate and from direct convergence between the Pacific and NorthAmerica plates.

McCrory, P. A., 1989, Late Neogene geohistory analysis of the Humboldt basin and its relationship to convergence of the Juan de Fuca plate: Journal of Geophysical Research, v. 94, p. 3126-3138.McCrory, P. A., 1995, Evolution of a trench-slope basin within the Cascadia subduction margin: the Neogene Humboldt Basin, California: Sedimentology, v. 42, p. 223-247.McCrory, P.A., 1996, Evaluation of Fault Hazards, Northern Coastal California: U.S. Geological Survey Open-File Report 96-656, 1 map sheet, 1:500,000, 87 p.McCrory, P. A., Wilson, D. S., and Murray, M. H., 1995, Modern plate motions in the Mendocino Triple Junction region: Implications for partitioning of strain (abs.): Eos, Transactions, American Geophysical Union, v. 76, n. 46, p. F627.Murray M. H. and Lisowski, M., 1995, Crustal deformation near the Mendocino triple junction (abs.): Eos, Transactions, American Geophysical Union, v. 76, n. 46, p. F627.Wilson, D. S., 1993, Confidence intervals for motion and deformation of the Juan de Fuca

plate: Journal of Geophysical Research, v. 98, p. 16,053-16,071.


74

The Transition FROM Subduction TO Transform Regime: Structural ANDStratigraphic Signatures Associated WITH Triple Junction Migration OffshoreNorthern California

Anne Meltzer and Sean Gulick

Dept. of Earth & Environmental Science, 31 Williams Dr., Lehigh University, Bethlehem, PA 18015,[email protected]

In northern California, the North American, Gorda, and Pacific plates come in contact forming theMendocino Triple Junction. The triple junction itself is a broad zone extending over several hundred squarekilometers laterally and includes the crust and upper mantle at depth. North of the triple junction the Gordaplate is subducting beneath North America forming the southern end of Cascadia subduction zone. South of thetriple junction, the Pacific and North American plates form a transform boundary with strike-slip motionoccurring along a series of faults associated with the San Andreas Fault system. West of the triple junction, thePacific and Gorda plates form a transform boundary along the Mendocino Fracture Zone. Seismicity andstructural and stratigraphic data indicate that the triple junction is centered onshore near Petrolia, however, thetectonic processes and strain related to plate boundary interaction radiate outward and influence crustaldeformation of coastal California between Clear Lake and Crescent City and extend offshore well west of thedeformation front of the Cascadia subduction zone.

The canonical view is that the triple junction has been steadily marching northward along the westernmargin of North America since the Pacific-Farallon spreading ridge intersected the margin approximately 28million years ago. As the triple junction migrates northward arc magmatism shuts off and convergence yields tostrike-slip with local extension and shortening accommodating any oblique component of motion between thePacific and North American plates. In the wake of the triple junction, a slab-free zone develops where there is nooceanic plate subducting beneath the margin and hot asthenosphere wells up toward the base of the crust.Thermo-mechanical models incorporating spatial and temporal heating of the crust due to the asthenosphericwindow predict localized weakening in the lower crust leading to deformation related to transform tectonics inthe upper crust.

In detail, the transition from subduction to transform motion has been complicated by the irregularboundary between the Pacific and Farallon plates, a series of ridge segments offset by fracture zones, thatprogressively interacted with the margin of western North America. A complex pattern of magnetic anomaliesoffshore California tell a tale of slowed spreading, plate fragmentation, stalled subduction, and eventual micro-plate capture as young, hot, negatively buoyant oceanic slabs east of the spreading ridge stall beneath theoverriding plate then are subsequently ceded to the Pacific plate. As this happens, the main locus of strike-slipshifts eastward farther into the North American continent and transform motion propagates rapidly northwardlengthening the transform margin on relatively short time scales.

A series of seismic reflection profiles from the Mendocino Triple Junction seismic experiment augmentedwith industry data sets provides an interesting view of the structural and stratigraphic development offshoreNorthern California during this period and provide a current snapshot capturing the plate boundaries in transitionfrom subduction to transform motion.

The Vizcaino block and overlying Neogene Pt. Arena basin sit immediately south of the triple junctionbetween the Mendocino and Pioneer fracture zones. The Vizcaino block is bound on the east by a well-defined 3-5 km wide shear zone mapped as the San Andreas Fault zone. The western boundary is the Oconostota Ridgewhich is underlain by 25-26 Ma Pacific oceanic crust. In the most simple plate model, this entire region hasbeen passively riding above the Pacific plate for its entire history, its northeastern corner coinciding with thetriple junction. In this position the Vizcaino block would have been shielded from deformation taking placenorth of the migrating triple junction and would sit west of the main transform plate boundary. Whileinterpretation of the reflection data is generally consistent with this model, the Vizcaino block can be brokeninto three distinct structural domains based on the character of basement reflections and thickness/deformation ofthe overlying Neogene sedimentary section. The region shows a remarkable amount of young deformation


75

(shortening) well south of the triple junction and west of the main locus of strike-slip associated with thetransform plate boundary. An older, high angle, through going fault is observed west of the shear zoneassociated with the San Andreas Fault. This fault offsets basement ~1.5-2 km in places (up to the east) anddeforms overlying sediments as young as Pliocene. The fault merges with the San Andreas approximately 50km south of Pt. Delgada. The shear zone associated with the San Andreas trends onshore at Pt. Delgada. If theSan Andreas curves northwest and trends offshore again as suggested in previous maps, it does so within thethree-mile state waters. There is no evidence of the fault trending offshore north of Pt. Delgada within theavailable seismic data.

Directly northwest of Pt. Arena, the southern end of the offshore Pt. Arena basin is characterized by a thick(>3.0 km), highly deformed sedimentary section including fault cored anticlines and anticline/syncline pairsabove faulted/folded basement. At least 2 distinct phases of deformation are observed. An early phase (pre 5 Ma)and a younger, predominantly Plio-Pleistocene phase which continues today, in places folding sediments nearthe seafloor on the western edge of the basin. Upper Miocene and younger sediments were largely funneled intolocal depocenters between growing anticlines and syntectonic deformation of these sediments dates fold growth.A major erosional unconformity separates the deformed sediments from westward prograding Pleistocene andyounger sediments. North of this region the degree of deformation diminishes. Sediments and basement are stillfolded and faulted, but the folds are broader and less frequent. In the far northern end of the basin, just south ofthe Mendocino Ridge, strong, coherent, flat to gently northward dipping basement reflections are observed. Athick, essentially undeformed sedimentary section overlies basement. Sediments are thickest immediately southand west of the triple junction (northeast corner of the basin). The Pt. Arena basin terminates at the Mendocinoridge. Sediments on the flank of the ridge are uplifted and tilted southward. There is a strong angularunconformity at the seafloor indicating the ridge crest was at one time at sea-level and has subsided over 1200meters.

The Mendocino Fracture Zone passes on the north side of the Mendocino Ridge. In this region theMendocino Transform Fault juxtaposes 25-26 my old oceanic crust of the Pacific Plate to the south against 6-7my old oceanic crust of the Gorda Plate to the north. Structural development of this region is affected by the agetransition across the Mendocino Fracture Zone; the young, hot, Gorda plate absorbs most of the deformation.The Eel River forearc basin overlies the subduction zone complex north of the Mendocino Fracture Zone. Thesouthern end of the Eel River basin is extremely deformed, shortened, uplifted, and in the process of beingcannibalized. The Southern end of the forearc lies beneath only 100 m of water and folded and faulted sedimentare truncated at the seafloor. The presence of the triple junction in the Eel River basin is currently felt at least60 km to the north as a broad region of uplift. As a result of this uplift, Upper Pleistocene and youngersediments bypass the southern end of the basin and are either deposited in a depocenter on the lower shelf or aretransported through the Eel Canyon to be deposited along the toe of the accretionary prism. The northern portionof the basin shows primarily dip-slip low angle faulting on the western edge of the basin and high-angle faultssuggestive of strike-slip deformation beneath the continental shelf. There is an intervening little deformed 30 kmwide marginal plateau (the Klamath Plateau), and the main forearc basin is in essence a single largesynclinorium. In the Southern portion of the basin, low-angle primarily dip-slip faults on the western edge ofthe forearc curve east-southeast towards the triple junction and change dip to become high-angle faults thatresemble positive flower structures suggestive of youthful strike-slip motion. Near the triple junction theoriginal forearc basin is subdivided into a series of smaller sub-basins.

From the toe of slope to the continental shelf deformation changes from convergence related low-anglethrusts to near-vertical faults suggesting a component of strike-slip. As the triple junction moves northward,faults originally accommodating shortening in the convergent plate setting north of the triple junction, rotateand steepen, serving as sites for strike-slip motion during and after passage of the triple junction. Strike-slipactivity on these eastern faults suggests that the transition to transform motion associated with the Pacific-NorthAmerican plate boundary is propagating northward.


76

A gravity study of the Puerto Rico Trench and surrounding region:Implications for the origin and tectonics of the Puerto Rico Trench

A. Miller1, N.R. Grindlay1, P. Mann2 and J. Dolan3

1Department of Earth Sciences, University of North Carolina at Wilmington, 601 S. College Road,Wilmington, NC 284032Institute for Geophysics, University of Texas at Austin, 4412 Spicewood Springs Road, Bldg. 600, AustinTX, 787593Department of Earth Sciences, University of Southern California, University Park, Los Angeles, CA 90089-0740

The worlds’ large free-air a gravity minima (-380 mGal) and the Atlantic Ocean’s deepestlocality are found at the Puerto Rico Trench (8450m). The trench lies to the north of the islandof Puerto Rico within a wide area of deformation associated with the oblique convergence ofthe North American plate beneath the Caribbean plate. Motion along this plate boundary isdominantly left-lateral strike slip. A puzzling feature of the Puerto Rico Trench gravityanomaly is that there are other deep trenches and oblique convergent plate boundaries aroundthe world that do not exhibit such a large amplitude anomaly. Several tectonic models havebeen proposed to explain the origin of the trench, it’s associated gravity anomaly and the rapidsubsidence of the Pliocene shallow-water carbonate platform on the northern insular margin ofPuerto Rico about 3.3 Ma. These models include: 1) extreme tectonic erosion associated withthe oblique underthrusting of aseismic ridges (Birch, 1986), 2) regional transtension (Speedand Larue, 1991), 3) and the favored model, loading of the North American by the Caribbeanplate, presently subducting northward along the Muertos Trough, beneath the island of PuertoRico (Dillon et al., 1996, Dolan et al., 1998) Recently collected bathymetry and single-channel seismic data together with existing multi-channel seismic data have been used determine realistic crustal and upper mantle structure forthe region. Two-dimensional forward modeling of gravity anomalies associated with the plateboundary have been performed in an effort to test the viability of these proposed models.Three roughly N-S trending profiles across the eastern, central and western portions of thetrench and Puerto Rico/ Virgin Islands arc have been modeled. Only the easternmost modelprofile that crosses the Virgin Islands platform appears to closely to match the observed data.Modeled profiles over the central and westernmost portions of the trench predict higher gravityvalues than observed. Since the 2-D models assume isostatic equilibrium, the differencesbetween modeled and observed data can best be explained by a dynamic component. Loadingof the North American plate at depth by the Caribbean plate, as proposed by Dillon et al.,(1996) would cause the plate to flex downward thus introducing a dynamic component thatwould cause the region not to be in isostatic equilibrium. Future plans include testing thishypothesis by modeling the expected flexural shape of a loaded North American plate andcomparing that with the shape of the Puerto Rico trench.

ReferencesBirch, F.S., Isostaic, thermal, and flexural models of the subsidence of the north coast of

Puerto Rico, Geology, 14, 427-429, 1986.Dillon, W., N. Edgar, K. Scanlon, and D. Coleman, A review of tectonic problems of the

strike-slip northern boundary of the Caribbean plate and examination by GLORIA, inGeology of the United States’ Seafloor: The View from Gloria, eds. J.V. Gradner, M.E.Field and D. Twichell, 135-164, 1996.


77

Dolan, J., H. Mullins, and D. Wald, Active tectonics of the north-central Caribbean: Obliquecollision, strain partitioning, and opposing subducted slabs, in Active Strike-Slip andCollisional Tectonics of the Northern Caribbean Plate Boundary Zone, eds. J. Dolan andP. Mann, GSA Special Paper 326,1-62, 1998.

Speed, R. and D. Larue, Extension and transtension in the plate boundary zone of thenortheastern Caribbean: Geophys. Res. Lett, 18, 573-576.


78

Three-dimensional structure of the Cordillera Oriental of Colombia: Asegment of the southern Caribbean plate boundary

Camilo Montes

306 G&G Building, Structural Geology and Tectonics, The University of Tennessee, Knoxville TN, 37996-1410 , fax: (423)974-9326, office phone: (423)974-0701, e-mail: [email protected]

Current palinspastic reconstructions of the Cordillera Oriental of Colombia assumedeformation paths perpendicular to the trend of the Cordillera (NW-SE dip-slip), ignoringstrike-slip deformation. Kinematic plate tectonic reconstructions (Ladd, 1976) predict that theCordillera Oriental of Colombia experienced oblique NE-SW convergence and transpressionthroughout the Cenozoic. Despite its regional setting, a doubly vergent fold-thrust beltgeometry (driven by subduction) has been suggested for the Cordillera Oriental, and about 150km of NW-SE horizontal shortening has been estimated using standard two-dimensionalreconstruction techniques. Deformation in the Cordillera Oriental must be related to transpression produced by obliqueconvergence, and dextral motion of the Caribbean Plate along the NW corner of South Americasince the Eocene. The overall structure of the Cordillera Oriental might resemble a dextral"palm tree"structure (Lowell, 1972; Sylvester, 1988) where the double vergence is theexpression of a room problem created as two blocks obliquely converge, and the material inbetween is thrust up. These upthrusts may flatten upwards and cause passive uplift, and milddeformation on the thick layer of Upper Cretaceous and Tertiary sediments along the axis of theCordillera. Oblique convergence may have been accommodated by thrust sheets moving out ofthe NW-SE plane usually chosen to model the two-dimensional structure. Therefore,prominent strike-slip faulting should not necessarily be present. We complied a comprehensive geologic database to test competing hypotheses regarding thetectonic evolution of the Cordillera Oriental (subduction vs. transpression). This databaseserved as starting point to build a three-dimensional computer model where forward modelingwas performed. This model illustrates the viability of transpression, instead of subduction, asthe driving mechanism for deformation of the northern Andes. Although this dataset is limited,and some of the underlying assumptions are open to question, this model highlights the needand the viability to study the structure of the Cordillera in three dimensions. Currently, fieldstudies are being carried out in the western flank of the Cordillera with the goal of measuringthe relative contributions of strike- and dip-slip to deformation.


79

Tectonics of the Ibero-African margin.

Teresa Moreno & Wes Gibbons

Department of Earth Sciences. Cardiff University. Cardiff CF1 3YE, Wales, UK.e-mail: [email protected]

The Iberian Peninsula lies at the western end of the Alpine-Himalayan orogen, caught between the twocontinents of Africa and Europe, the widening Atlantic Ocean, and Neotethyan subduction systems in thewestern Mediterranean. Iberian tectonism over the last 100Ma reflects this unique position, with deformationoccurring on both the northern (Pyrenees) and southern (Betic) margins of the Peninsula as it moved broadlyeastward and rotated anticlockwise, sometimes as an independent plate, sometimes linked to Africa or Europe.Since early Miocene times (c.24Ma) Alpine deformation in Iberia has focused along the southern margin whichbecame the African-European plate boundary. The late Mesozoic to Recent tectonic history of the Betic orogenprovides a record of Ibero-African interactions, but the geology is complex and controversial. Successive eventshave interleaved mantle lherzolites with crustal rocks, emplaced blueschists over the Iberian margin, and inducedmassive extensional collapse which was partly synchronous with thrusting, strike-slip faulting, oroclinalrotations, and calc-alkaline magmatism. During the Early Cretaceous Iberia, mainly linked to Africa, had rotated 30o anticlockwise with respect toEurope. Continuing eastward drift during the mid-Cretaceous was achieved independently of both Africa andEurope, with the Azores-Albor·n tectonic line (AATL) acting as a southern transform plate boundary. LateCretaceous Iberian movements once again closely paralleled those of northwest Africa, and movement on theAATL declined. Thick, deep marine, late Mesozoic to Palaeogene siliciclastic sediments (later to become theFlysch Nappes) were deposited in transcurrent basins between Iberia and Africa. Mesozoic transtensionalthinning and upwelling of lower lithospheric continental mantle along the Ibero-African margin has beeninvoked to help explain the later introduction of fertile lherzolite to the Tertiary Betic nappe stack. Africa andIberia continued to move ENE across the K-T boundary and during the early Palaeogene. These movements wereaccommodated further east by the subduction of oceanic crust in the western Mediterranean. The late Cretaceousto Palaeogene blueschists and eclogites produced by this subduction were destined to be stacked over thecontinental basement of Europe and are currently exposed in the Western Alps, Corsica, and the Betics. As the rotation pole of Africa moved progressively northwards during the Eocene and Oligocene, Iberiaincreasingly once again behaved as a separate plate, bounded by the Pyrenees and the Azores-Albor·n tectonicline (AATL). Africa moved ENE with respect to Europe, sliding transpressionally against Iberia which alsocontinued to drift eastward. HP-LT accretionary prism rocks were thrust over continental basement as thewestern Mediterranean subduction system collided with western Europe. Most authors envisage a SE-dippingslab (as shown here: this is more in accord with Corsican geology), although a NW dip has been suggested. Theage of internal Betic nappe stacking is poorly constrained but generally attributed to the Palaeogene. Morecontroversial is the timing of lherzolite exhumation: although one of the largest known masses of orogenicfertile mantle rocks, whether they were exhumed after oblique Palaeogene collision or by Neogene extensionremains unproven. Also uncertain is the number of Palaeogene deformation events recorded in the Betics, withsome authors preferring alternating compressional and extensional phases rather than only one majorcompression. A major problem inhibiting understanding of Palaeogene events is a later widespread overprint ofNeogene extension. In early Miocene times Iberia became attached to Europe as Pyrenean movements ceased and Africa continuedits ENE drift. In the previously thrust-stacked internal zones of the Betic orogen, lithospheric extensioncommenced and led to the (Aquitanian) onset of sedimentation in the Alboran Sea. The Betic Cordillera becamean archipelago similar to the modern Aegean Sea. Regional extension ran NE-SW from the Provencal-Ligurianbasin NW of Corsica through the Valencian Trough to the Alboran Sea, transecting the earlier arc-continentcollisional front and migrating eastward with time. The extension in Corso-Sardinia and Italy was, and in placesstill is, associated with calc-alkaline to shoshonitic magmatism, and has been linked to SE retreat of the modern


80

Calabrian subduction zone. Lower Miocene magmatism in SE Iberia began in the western Betics, initiallyproducing arc tholeiite dyke swarms (Aquitanian), then became more organised by the mid-Miocene when majorandesitic caldera complexes erupted in the eastern Alboran Sea. Volcanism in the eastern Alboran regioncontinued into the late Neogene, locally erupting more evolved (dacite to rhyolite) magmas from calderas, andbecoming generally more K-enriched with time to produce Late Miocene to Pliocene shoshonites, alkali basalts,and lamproites. Currently competing hypotheses presented to explain the dramatic Miocene extension in the Alboran Sea areaemphasise one of the following models:1 Convective removal of a thickened lithospheric root: initial rise of hot asthenosphere produces uplift at the

Earth’s surface followed by collapse to produce a basin with a fixed depocentre. The Betics are commonlyviewed as a world classic example of this kind of radially directed “extensional collapse” as vertical stressescaused by elevated crust surpass the horizontal stresses induced by plate convergence.

2 SE-directed delamination of thickened subcrustal lithosphere: initial rise of hot asthenosphere produces uplift atthe Earth’s surface followed by extension as the lithosphere peels away and moves laterally to produce abasin with a migrating depocentre. This model is used to explain the eastward migration of sedimentation inthe Alboran Sea.

3 Collision with NW-dipping subduction zone then slab detachment: Palaeogene oblique convergence betweenIberia and a NW-dipping Mediterranean subduction zone is followed by Neogene detachment and rapidsinking of the lithospheric slab. Release of slab-pull force beneath the thickened Betic orogen, combinedwith lateral inflow of hot asthenosphere, induces uplift and subsequent extension, with elevated isothermsand decompression melting driving calc-alkaline volcanism.

4 Collision with SE-dipping subduction zone then polarity reversal and slab rollback: Late Mesozoic- Palaeogeneconvergence between Europe and a SE-dipping subduction zone produces collision from the W Alps to theBetics. A new, late Oligocene to Quaternary, orogenic system with an opposite sense of polarity then cutsobliquely through the earlier thrust-stacked orogen. NW-dipping subduction retreats SE to induce intra-arcextension from Provence to the Alboran Sea. This is also consistent with eastward migration ofsedimentation in the Alboran Sea.

Convergence along the Ibero-African margin became more compressive in the mid-Miocene (16Ma) whenAfrica turned northwards and initiated the transcurrent trans-Alboran/Eastern Betics shear zone running fromMorocco through SE Spain. Another anticlockwise plate rotation took place in the late Miocene, turning Africamore northwestwards into Iberia and reducing the size of the Alboran Sea. A similar intensification of NW-directed convergence in the late Pliocene induced further uplift and inversion of late Neogene sedimentary basins.A return towards more NNW-directed convergence during the early to mid-Pleistocene reorganised fault and basinkinematics and began the development of the present day landscape. Many Quaternary Ibero-African movementshave focussed around the sigmoidal Eastern Betics shear zone which is transpressive in the NE (Alhama deMurcia fault system), transcurrent in the centre (Palomares fault system) and transtensional in the SE(Carboneras fault system). This tectonic lineament is the modern expression of a long-active Neogene strike-slipboundary separating a transtensional western Alboran sector from a transpressive eastern Alboran sector. Theboundary probably roots into a low-angle mid-crustal detachment beneath the Betics (8-10km depth) that haspresumably been reactivated many times since Palaeogene collision. Combined deep (>600km) earthquake datain southern Spain and seismic tomography studies indicate the current presence of a vertical slab of seismicallyfaster material between 200-700km below the Betics and Alboran Sea. Other modern seismic activity occurs atshallow or intermediate (<180km) depths and is attributed to current oblique convergence between Africa andEurope. Thus the late Mesozoic-Cenozoic tectonic evolution of the Betic orogen reflects interactions between twoMediterranean subduction events and oblique Ibero-African plate movements. Neogene extension in the AlboranSea was probably mainly, or wholly, driven by subduction processes rather than convective removal ordelamination of a thickened lithospheric root. This interpretation is the most consistent with an overview ofMediterranean tectonics and especially emphasises: 1. The linkage in space and time between the Alboran Seaand the opening of the suprasubduction Ligurian-Valencian basins. 2. Calc-alkaline to high-K magmatismthroughout Neogene time in the Alboran region: comparison is drawn with the still-active volcanicity in theCalabrian arc of S Italy. 3. The presence of a dense vertical slab at depth beneath the Betics and Alboran sea: thisis interpreted as a detached remnant of subducted oceanic lithosphere.


Results from the Neotectonics studies in Western Puerto Rico.

Juan Carlos Moya1.

13114 Broadway # 3. Boulder, CO 80304. [email protected]

Puerto Rico is located in a tectonic environment associated with the oblique subduction of abuoyant marine ridge, the Bahama Bank. A reconstruction of the tectonic history of Puerto Ricoshows that the effects of this oblique subduction have been complex. In the Eocene to the MiddleOligocene, compression created a suture zone between island blocks and strike-slip faults were alsodeveloped in the island. The Late Oligocene to Middle Miocene was a period of quiescence. Afterthis period, from the Late Miocene to the Pliocene, the passage of the Bahama Bank underneath thePuerto Rico Microplate induced the island to rotate. It is suggested that during this time, from theMiocene through the Pliocene, Puerto Rico and eastern and central Hispaniola were part of thesame microplate. Rotation of this microplate around a single rotation pole, located in the CaribbeanPlate south of western Puerto Rico, generated the opening of the Anegada Passage together with azone of extension in southern Puerto Rico, the formation of the Muertos Trough, and a zone ofcompression in central Hispaniola. Geomorphologic, geologic, and structural data suggest thatsomewhere in the Pliocene maybe Late Pliocene and Early Quaternary, southwestern Puerto Ricopresented an environment of strike-slip faults with transtensive and transpressive faults.Information from two faults located in western Puerto Rico, the Cordillera-Mayaguez Fault Systemand the Joyuda Fault, suggest that after the rotation, Plio-Quaternary strike-slip fault movementsthe regional geomorphology of western Puerto Rico. A minimum of 50 m of uplifting has beensuggested for southwestern Puerto Rico. Recent GPS vectors show that Puerto Rico and Hispaniola are now moving apart at about 4-5mm/yr. This separation has created an extensional seismotectonic regime affecting the easternportion of Hispaniola, the Mona Passage, and western Puerto Rico, within a boundary zonebetween the North American and Caribbean Plates. The zone of extension between Puerto Ricoand Hispaniola suggested by the GPS data can be explained by the recent attachment of PuertoRico to the Caribbean Plate, which is moving to the east. Thus, the former Puerto Rico microplatehas partitioned into two smaller microplates; the new Puerto Rico microplate (Puerto Rico) and theEl Seibo microplate (eastern Hispaniola). Their separation has been generated by the release of thePuerto Rico microplate by the Bahama Bank and the recent locking of Hispaniola by this buoyantfeature. The zone of extension then includes eastern Hispaniola, the Mona Passage and WesternPuerto Rico (Figure 1). This extension can be explained as a new seismotectonic regime forWestern Puerto Rico which changed from strike-slip to extension. Paleoseismic information on tsunami deposits and paleoliquefaction show the importance of theseismic hazards in western Puerto Rico associated with this extension. Some of the normal faultsas suggested by recent focal mechanisms, could be reactivated faults, which were originally strike-slip faults. These normal faults and the others in the Mona Passage and Eastern Hispaniolarepresent an important seismic risk for western Puerto Rico having potential magnitudes up toM7.5. The recent paleoseismic information on tsunami deposits and on liquefaction injectionssuggests that there was one important event between 1300 and 1511. The most recent eventgenerating tsunami deposits was in 1918. The 1300-1500 event is the first pre-historic eventdefined in western Puerto Rico. Many areas in Western Puerto Rico present soft deposits with highpotential to some earthquake induced hazards such as: liquefaction, ground motion amplification,lateral spreads, landslides, and tsunamis.


Figure 1. Areas under extension between Puerto Rico, Mona Passage, and Eastern DominicanRepublic showing some important faults as suggested by structural, geomorphologic, seismic, andmarine data.

19o18.5o18o68.5o67.5o66.5oPuerto RicoDominican Republic MonaCanyon MonaPassageCaribbean SeaAtlantic Ocean CMFSGSPRFZFaultsZone underExtension?????????


83

Transpressional History of the Serrania del Interior Foreland Fold and ThrustBelt, North-Central Venezuela: New Evidence from Apatite Fission-TrackModeling and 2D Seismic Reflection Analysis.

Jaime Perez De Armas

Rice University, 6100 South Main St., Department of Geology and Geophysics, MS-126, Houston, TX 77005-1892, USA.

The Caribbean-South American plate boundary along north-central Venezuela is a megasuturewhich consists of three east-west trending belts: 1) Cretaceous metavolcanic rocks of the Dutchand Venezuelan Leeward Antilles, 2) Jurassic-Cretaceous meta-sedimentary and metavolcanicrocks of the Caribbean Mountain System, and 3) Mesozoic-Cenozoic sedimentary rocks of theSerrania del Interior foreland fold and thrust belt and Guarico sub-basin. Deformationalstructures in the plate boundary zone indicate partitioning of the right-oblique convergence-ratevector into two components resulting in south-vergent thrusting and dextral displacementsalong E-W trending faults. Within the Caribbean Mountain System, the post-Miocene LaVictoria fault zone is the southernmost expression of dextral strike-slip motion in continentalnorth-central Venezuela. Southward, deformation is taken up solely by the plate boundaryperpendicular shortening component. The western Serrania del Interior foreland fold and thrust belt and the Guarico sub-basin arethe southernmost units of the diffuse transpressional Caribbean-South American plateboundary. This fold and thrust belt involves Upper Cretaceous passive-margin deposits,Maastrichtian-Lower Eocene flysch, and Oligo-Miocene foredeep deposits. Within the Guaricobasin, the sequence consists of Paleozoic pre-rift and Jurassic syn-rift sedimentary rocks of theEspino graben, both of which rest unconformably either on Paleozoic metamorphic rocks or oncrystalline basement of the Guyana craton. Unconformably overlying the economic basement,are Cretaceous passive margin siliciclastic and carbonate rocks. The foredeep consists ofLower to Middle Oligocene transgressive litoral sandstones, Middle to Upper open-marineshales, and a regressive Miocene siliciclastic sequence. Fission-track modeling of Maastrichtian-Lower Eocene rocks from the Serrania suggests thattwo important cooling events occurred in the study area in 1) Middle Eocene and 2) LateOligocene-Early Miocene. These apatite fission track ages do not seem to fit the model ofyounging of Caribbean-South American plate collisional events from west to east during theCenozoic. The Middle Eocene cooling event has been found near the Ensenada de Barcelona.This Middle Eocene cooling event has also been observed in apatite fission-track analyses ofcore samples from the Lake Maracaibo basin. The Late Oligocene-Early Miocene fission-trackages occur in the Cretaceous-Paleogene rocks in the entire area. Fission-track modeling ofOligo-Miocene samples from the Serrania and the Guarico sub-basin suggests as well that theserocks were not heated above ~ 110 °C. Interpretation of seismic reflection profiles (1500 km) allows to constrain the tectonic modelfor the area. Four stratigraphic sequences have been recognized in the Guarico sub-basin.These sequences rest on top of an economic basement. The latter includes igneous-metamorphic basement of the Guyana craton and pre-Cretaceous pre-rift and syn-riftsedimentary deposits. The recognized sequences are: Sequence 1) Cretaceous transgressivesiliciclastic and carbonate rocks; Sequence 2) Lower Oligocene to Upper Oligocene-LowerMiocene foredeep deposits; Sequence 3) Miocene siliciclastic rocks, and Sequence 4) Mioceneand younger deposits. On the basis of apatite-fission track modeling and interpretation of seismic reflection profilesit is suggested that the tectonic evolution of the western Serrania del Interior and Guarico


84

foredeep involves seven phases of deformation: 1) Triassic-Jurassic inception of the EspinoGraben, 2) Generation of a Late Cretaceous-Early Eocene foredeep along northern Venezuela,3) Middle Eocene southward accretion and destruction of the Late Cretaceous-Early Eoceneforedeep, 4) Early Oligocene inception of the foredeep, 5) Early Oligocene NW-SE extensionin the basin, 6) Neogene thrusting, 7) Pliocene to Recent regional uplift. The Paleozoic and Cretaceous autochthonous sedimentary rocks and non-decoupled LowerOligocene foredeep sequences may extend below the Serrania del Interior and possiblyunderneath the accreted terranes to the north.


85

DIAPIR-FED MUDFLOWS AND SERPENTINITES: STRIKE-SLIPRECYCLING OF BUOYANT MATERIAL IN HISPANIOLA’S OBLIQUE-SLIPFOREARC

James Pindell

Dept. Earth Sci, Oxford U., Oxford OX1 3PR England; also Dept. Geology, Rice University, Houston, Tx.

INTRODUCTION AND REGIONAL SETTINGIn northern Dominican Republic, the fault-bounded (10-30km) “Puerto Plata Inlier” (PPI) is flanked by

the coastline (north), the Camu Fault and Neogene carbonate rocks (south), and Neogene carbonate rocks (eastand west). The PPI exposes rocks belonging to a Cretaceous-Eocene forearc/subduction complex from beneath aNeogene carbonate cover (see Fig., field guide). The linear E-W Camu Ft bounding the PPI’s south side issinistral and stratigraphically N-side up at the PPI contact, but the Neogene carbonates on the fault’s S flank sithigher (C. Septentrional), suggesting relatively faster erosion within the PPI. Projecting NE from Camu Ft andoccurring within in PPI, the alluvial Maimon Valley separates two structurally distinct parts of the PPI. To thewest, Basement Complex rocks form a morphologically mature landscape, whilst equivalent/other rocks east ofthe Valley comprise a relatively high relief, immature morphology.

This morphological difference indicates recent/active diapirism of relatively buoyant/mobile lithologiesand mélanges of the subduction complex east of the Valley. This is further suggested by the occurrence of theFranciscan-like San Marcos mud-diapir/flow unit which extrudes and “flows” N-ward across all strata in the PPIfrom a portion of Camu Ft which was/is a conduit for mud diapirs. Diapirism, or faulting associated with it, hasalso disrupted the formerly continuous Neogene carbonate cover.

The isolated occurrence of Neogene carbonates of Pico Isabela, which rests on the Basement Complex,appears to have been tectonically dislocated from the Neogene carbonate section outside the PPI. The buoyantand mobile San Marcos unit (and several serpentinite bodies) fills the void created by this dislocation (see Fig.,field guide). Field relations prevent a clear understanding of that dislocation, but a half graben (master E-dippingfault surfacing at Maimon Bay “graben”?) has probably formed along the north side of the sinistral Camu Ft (seeFig., field guide), suggesting extensional breakup within the PPI.

Although San Marcos derives from the Cretaceous-Eocene subduction complex below, it has alsopicked up knockers as young as Middle Miocene during diapirism and subsequent sub-horizontal gravitationalflow. The unit today maps as a Late Miocene-Pleistocene lobate wedge. This 2-part history has causedconfusion, namely that San Marcos might represent a Miocene subduction mélange (e.g., Bourgois et al.,1982).

GEOLOGY AND GEOLOGICAL HISTORYPuerto Plata Basement Complex (PPBC) of the PPI comprises Lower Cretaceous-Eocene serpentinites,

tectonized harzburgite, cumulate gabbroic rock, and mafic/intermediate volcanic rock sometimes with redinterpillow cherts, interpreted as remnants of an ophiolite. The non-Tethyan cherts (Montgomery, 1992/4)suggest pieces of far-travelled Caribbean forearc rather than obducted slivers of Proto-Caribbean (Tethyan).

The PPBC is overlain by two units of similar age: 1, Paleocene-Early Eocene Imbert Fm off-whitecrystal tuffs, vari-colored cherts, and sandy to pebbly turbiditic rocks. Coarse clastic beds are more common lowin this formation, and contain sand and pebbles of serpentinite, volcanic and metamorphic rock, and limestone;and 2, sedimentary serpentinite conglomerates and algal-limestone buildups or patch reefs of the ?earlyPaleogene shallow-water La Isla Fm. La Isla limestone buildups and serpentinite conglomerates containfragments of one another, and corals were observed in growth attachment to serpentinite cobbles, indicatingcoeval, shallow-water deposition of the conglomerate and limestone. Imbert Fm is believed to be older than LaIsla because it is more deformed and because Imbert deposition occurred in deeper water than La Isla deposition,possibly prior to or during uplift from forearc platform to photic water depths. The period of La Isla depositionand probably longer thereafter was a time of erosion across much of the PPI, leading to an angular unconformitybetween Imbert and younger units. The Eocene uplift and erosional hiatus is interpreted as recordingconvergence/collision between Greater Antillesarc and Bahamas.


86

In Late Eocene-Oligocene, shelf deposition is indicated by conglomerates and terrigenous, mica-bearingsands/marine shales of Luperon Fm. Tectonism at this time was apparently minor. Deformation and erosion(emergence?) occurred again in Early Miocene, but deposition was renewed in Mid to Late Miocene with VillaTrina Fm carbonates (terrigenous input reduced). Early Miocene deformation/change in sedimentation maypertain to strike-slip separation of Hispaniola from SE Cuba along Oriente Transform Ft.

Regional uplift occurred again since Pliocene, elevating the Pliocene reefal cap of Villa Trina Fm toseveral hundred meters, and exposing ?Pleistocene reefal limestones/beachrock along the coast. This uplift is dueto oblique convergence of Hispaniola with easternmost Bahamas Bank along the Oriente transform. Finally,since ?Pliocene and probably associated with the Pliocene-Recent uplift, the San Marcos mud unit carrying mmto 100 m sized blocks of all the above mentioned lithologies (and others) diapirically rose to the surface andflowed horizontally across areas of the Imbert, Luperon and Villa Trina Fms.

THE CAMU FAULT AND MAIMON VALLEYCamu Ft defines the PPI’s southern limit. Fault gouge, breccias and steep dips in Villa Trina Fm (with

sinistral indicators) are common in the fault zone. No definitive strike-slip offset markers are known, but thePPBC may correlate with Río San Juan Complex across Camu Ft to the east, possibly suggesting 50-60 km ofmainly late Neogene displacement. Villa Trina Fm of C. Septentrional is topographically higher than the PPInorth of Camu Ft. but along the PPI’s length the Camu juxtaposes older (north) with younger rocks (south).Thus, the Camu is post-Miocene north side up as well here. The north side's relative uplift is superposed uponthe regional uplift of C. Septentrional. The fault's linearity for ~100 km suggests a steep dip near the surface.Vertical displacement is up to 300m (thickness of Villa Trina flanking the PPI).

Maimon Valley is also fault controlled, a graben along an E-W sinistral shear (Mann et al. 1984).Vertical offset is ~200 m as indicated by elevation differences of La Isla Fm outcrops within vs to the west ofMaimon Valley. Fault breccias occur along the valley's W flank and on the S-ward extension of the east flank.The E flank separates immature, diapirically-generated topography to the east, from more mature morphologiesto the west. Maimon Valley defines the western limit of most active tectonism in PPI.

The east flank of the valley appears to be the hanging wall of a NE-trending half graben whose masterfault dips E from the trace of Maimon Valley. Further, the diapirically-generated topography within this hangingwall may extrude along collapse faults within the hanging wall, thereby creating locally high, immature relief.North of Imbert, and east of the west boundary fault of Maimon Valley, the Luperon and Imbert formations diphomoclinally west, possibly indicating rotation of the half-graben’s hanging wall.

SAN MARCOS UNITSan Marcos unit occurs in 2 or 3 N-projecting lobes from Camu Ft and comprises blocks of the PPBC

plus metamorphic rocks (marbles, greenschists, blueschists, amphibolites) chaotically contained within a non-lithified sheared mud matrix with Eocene fauna (Nagle, 1979). Middle Miocene Villa Trina marls/limes are alsocommon, indicating a post-Middle Miocene age for its emplacement or deposition, even though the matrix isdominantly composed of Eocene material (Nagle, 1966). Mud diapirism and subsequent lateral flow (post-Miocene) is responsible for San Marcos unit as it maps today, as suggested by: 1, San Marcos muds areassociated with brecciated serpentinites, locally intruding as well as resting upon the serpentinites and otherPPBC constituents; 2, diapirism provides a mechanism for emplacing San Marcos in shallow water and/orsubaereal conditions; 3, diapirism/lateral flow explains N-wardly thinning lobes of San Marcos unit; 4,diapirism explains the pervasive shear stria, or slickensides, of the matrix. Diapirism was post- Villa Trinatime, as there is no interstratification or contamination between/by the two, and probably relates to post-Miocene tectonism on Camu and Maimon faults, and regional uplift of C. Septentrional. San Marcos matrixsources from Imbert Fm tuffs plus PPBC serpentines (Nagle, 1966). The Eocene San Marcos fauna are Imbertequivalent, but do not date final emplacement. San Marcos blocks may have been incorporated into the mobilematrix either during subduction or during diapirism and lateral flow.

IMPLICATIONS FOR FOREARC GEOLOGYThe ability of buoyant lithologies and mélanges to mobilize diapirically to the surface and, in so doing,

to incorporate cover rocks that are far younger than the true age of subduction, is a process which must beconsidered when interpreting the period of subduction from the rock assemblages of subduction mélanges.Further, such reworking of buoyant materials should be expected especially in zones of highly obliqueconvergence, where the strike-slip component of motion often occurs at trench-parallel faults (such as Camu Ft)which can easily serve as migration conduits for buoyant materials from depth to the surface.

FIGURES AND REFERENCES (see field guide)


87

Paleoseismic studies along the Septentrional fault, Dominican Republic

Carol S. Prentice1, Paul Mann2, Luis R. Peña3, G. Burr4

1U.S. Geological Survey, 345 Middlefield Rd., MS 977, Menlo Park, CA 940252 Institute for Geophysics, 4412 Spicewood Springs Road, University of Texas, Austin, TX, 787133Avenida Cuesta Colorado Esquina Calle no. 9, Las Colinas, Santiago, Dominican Republic4 NSF Accelerator Facility, University of Arizona, Tucson, AZ 85721

Along most of the North American-Caribbean plate boundary, active faults are offshore andunavailable for study using typical paleoseismic techniques. However, the Septentrional faultzone (SFZ) is the major, strike-slip, North American-Caribbean plate-boundary structure at thelongitude of Hispaniola, and it traverses the densely populated Cibao Valley of the DominicanRepublic, forming a prominent scarp in alluvium (Prentice et al., 1993; Mann et al., 1998).Study of the SFZ provides an opportunity to understand and quantify the late Quaternarybehavior of this part of the plate boundary, and the seismic hazard it represents for the islandsof Puerto Rico and Hispaniola. Recent GPS-based geodetic studies show high rates of motionbetween the North American and Caribbean plates (21 mm/yr) suggesting the potential for largeaccumulated elastic strain across the SFZ (Dixon et al., 1998). Our investigations show thatthe most recent ground-rupturing earthquake along this fault in the north-central DominicanRepublic occurred about 800 years ago, and involved a minimum of about 4 m of left-lateralslip. Our studies of offset stream terrace risers at two locations, Rio Juan Lopez and RioLicey, provide late Holocene slip-rate estimates of 6-9 mm/yr, and a maximum of 11-12mm/yr, respectively, across the Septentrional fault. Three excavations, two near Tenares andone at the Rio Licey site, yielded evidence for the occurrence of several prehistoricearthquakes. Dates of strata associated with the penultimate event suggest that it occurred post30 AD. These studies indicate that the SFZ has likely accumulated elastic strain equivalent to asignificant earthquake during the approximately 800 years since it last slipped, and should beconsidered likely to produce a destructive future earthquake. These data also indicate thatseveral large earthquakes in the historical record were not produced by the Septentrional fault,and most likely occurred within the offshore thrust belt described by Dolan et al. (1998).

References

Dixon, T. H., Farina, F., DeMets, C, Jansma, P., Mann, P., and Calais, E., 1998, Relativemotion between the Caribbean and North American plates and related boundary zonedeformation from a decade of GPS observations: Journal of Geophysical Research, v. 103,p. 15,157-15,182.

Dolan, J. F., Mullins, H. T., and Wald, D. J., 1998, Active tectonics of the north-centralCaribbean: Oblique collision, strain partitioning, and opposing subducted slabs, in: Dolan,J. F., and Mann, P., eds., Active Strike-slip and Collisional Tectonics of the NorthernCaribbean Plate Boundary Zone: Boulder, Colorado, Geological Society of AmericaSpecial Paper 326, p. 1-61.

Mann, P., Prentice, C. S., Burr, G., Peña, L. R., and Taylor, F. W., 1998, Tectonicgeomorphology and paleoseismology of the Septentrional fault system, DominicanRepublic, in: Dolan, J. F., and Mann, P., eds., Active Strike-slip and Collisional Tectonicsof the Northern Caribbean Plate Boundary Zone: Boulder, Colorado, Geological Society ofAmerica Special Paper 326, p. 63-123.

Prentice, C. S., Mann, P., Taylor, F. W., Burr, G., and Valastro, S., Jr., 1993,Paleoseismicity of the North America-Caribbean plate boundary (Septentrional fault),Dominican Republic: Geology, v. 21, p. 49-52.


88

Figure 1: Map showing locations of paleoseismic excavation sites along the Septentrional fault,Dominican Republic. Rectangle in inset shows area of larger map.


89

The northernmost San Andreas fault near Shelter Cove, California: New dataregarding an old debate

Carol S. Prentice1, Dorothy J. Merritts2, Edward C. Beutner2, Paul Bodin3, Allison Schill2*,Jordan R. Muller2**

1U.S. Geological Survey, 345 Middlefield Rd., MS 977, Menlo Park, CA 940252Geosciences Department, Franklin & Marshall College, Lancaster, PA 17604-30033CERI, Memphis State University, Memphis, TN* Present address: North American Publishing Co, 401 North Broad Street, Philadelphia, PA** Present address: Department of Geology and Geophysics, University of Hawaii, 1680 East-West Road,Honolulu, HI 96822

The location of the San Andreas fault in the Shelter Cove area of northern California hasbeen the subject of long-standing controversy within the geological community (e.g.,McLaughlin et al., 1983; Wald et al., 1993; McLaughlin et al., 1994; Brown, 1995). Surfacefaulting associated with the 1906 San Francisco earthquake was reported for a distance ofapproximately 435 km along the San Andreas fault (Lawson, 1908). For most of this distancethe fault is on land, but north of Point Arena the fault lies offshore. Near Shelter Cove, about120 km north of Point Arena, 1906 fault rupture was reported onshore for a distance of about 3km (Lawson, 1908). However, interpretation of the Shelter Cove surface rupture as acontinuation of 1906 slip along the San Andreas fault has been questioned in severalpublications that discuss results of regional geologic mapping (McLaughlin et al., 1983; 1994;Underwood et al., in press). McLaughlin et al. (1983, 1994) interpret their mapping to showno large-scale, post-middle Miocene offset across faults in the area where the 1906 ruptureswere observed. They propose instead that the San Andreas fault is either offshore or inlandfrom Shelter Cove, and that the observed 1906 ruptures were the result of shaking-related andgravity-driven deformation, rather than San Andreas fault slip. Other workers, however, interpret the 1906 reports, along with geomorphic and geologicfeatures in the area, as evidence of an onshore continuation of the San Andreas fault nearShelter Cove (Brown and Wolfe, 1972; Slosson, 1974; Brown, 1995; and Hart, 1996). Recentanalysis of geodetic data supports the interpretation that fault slip occurred this far north in1906 (Thatcher et al., 1997), but analysis of teleseismic waveforms in 1906 seismograms findsnothing that requires slip along the fault north of Point Arena (Wald et al., 1993). Wald et al.(1993) conclude that seismic data from the 1906 earthquake do not provide evidence to supportfault slip north of Point Arena. Our study (Prentice et al., in press) uses historical research, excavation across the 1906rupture, and independent geologic and geomorphic mapping to provide additional data bearingon whether or not a Quaternary fault is present onshore near Shelter Cove, and whether thesurface ruptures reported in the region in 1906 were the result of faulting. We conclude that thesurface ruptures reported in 1906 were the result of strike-slip faulting, and that a significant,Quaternary fault is located onshore near Shelter Cove. The apparent offset of Telegraph Creek(180 m), combined with a minimum radiocarbon age of about 13,000 years, suggest theHolocene slip rate of this fault is greater than about 14 mm/yr, indicating that it plays animportant role within the modern plate boundary system. The onshore trace of the fault zone iswell expressed as far north as Telegraph Hill, about 3 km north of Shelter Cove; north ofTelegraph Hill, its location is less well constrained, but we propose that a splay of the faultmay continue onshore northward for at least 9 km to the vicinity of Saddle Mountain.


90

ReferencesBrown, R. D., 1995, 1906 surface faulting on the San Andreas fault near Point Delgada,

California: Bulletin of the Seismological Society of America Bulletin of the SeismologicalSociety of America, v. 85, p. 100-110.

Brown, R. D., and Wolfe, E. W., 1972, Map showing recently active breaks along the SanAndreas fault between Point Delgada and Bolinas Bay, California: US Geological SurveyMiscellaneous Investigations Map, v. I-692, 1:24,000.

Hart, E. W., 1996, San Andreas fault Shelter Cove area, Humboldt County: CaliforniaDivision of Mines and Geology Fault Evaluation Report, v. FER-243.

Lawson, A. C., 1908, The California earthquake of April 18, 1906, Report of the StateEarthquake Investigation Commission, Carnegie Institute of Washington.

Wald, D. J., Kanamori, H., Helmberger, D. V., and Heaton, T. H., 1993, Source study ofthe 1906 San Francisco earthquake: Bulletin of the Seismological Society of America, v.83, p. 981-1019.

Thatcher, W., Marshall, G., and Lisowski, M., 1997, Resolution of fault Slip along the 470-km-long rupture of the great 1906 San Francisco earthquake: Journal of GeophysicalResearch, v. 102, p. 5353-5367.

McLaughlin, R. J., Lajoie, K. R., Sorg, D. H., Morrison, S. D., and Wolfe, J. A., 1983,Tectonic uplift of a middle Wisconsin marine platform near the Mendocino triple junction,California: Geology, v. 11, p. 35-39.

McLaughlin, R. J., Sliter, W. V., Fredericksen, N. O., Harbert, W. P., and McCulloch, D.S., 1994, Plate motions recorded in tectonostratigraphic terranes of the Franciscan complexand evolution of the Mendocino Triple junction, northwest California: U.S. GeologicalSurvey Bulletin, v. 1997, 60 p.

Prentice, C.S., Merritts, D.J., Bodin, P., Beutner, E., Schill, A., and Muller, J., in press,The northern San Andreas fault near Shelter Cove, California: Geological Society ofAmerica Bulletin.

Slosson, J. E., 1974, State of California Special Studies Zones Map, SE 1/4 Point DelgadaQuadrangle, California (1:24,000): California Division of Mines and Geology.

Underwood, M. B., Shelton, K. L., McLaughlin, R. J., Laughland, M. M. and Solomon, R.,in press, Middle Miocene paleotemperature anomalies near the San Andreas fault ofnorthern California: Geological Society of America Bulletin.

Figure 1: Map showing location of Shelter Cove and San Andreas fault.


Part 2: Slab break-off and lateral extrusion in the Eastern Alps and WesternCarpathians: Kinematics and analogue material modeling

Lothar Ratschbacher1, Blanka Sperner2

1 Geological Institute, University of Würzburg, Pleicherwall 1, 97070 Würzburg, Germany2 Geophysical Institute, University of Karlsruhe, Hertzstr. 16, 76187 Karlsruhe, Germany

Here we describe the collision - strike-slip subduction transition of the Alpine-Carpathian regionand explore the effects of a weakly confined lateral boundary, imposed by subduction, on theorogenic architecture of the Eastern Alps.

1. Why did the collision to subduction transition originate at all?East of the Alps, along which the European foreland has an ~ E-W trend, the foreland recedes,giving way to an embayment along which the i230° orogenic arc of the Carpathians was built(Fig. 2). About S-ward subduction and overall N-ward drift of Adria led to discontinuouscollision along the strike of the Alps-Carpathians and to the coupled system of continentalcollision along the Alps and subduction along the Carpathians.

2. Why did a strike-slip transition originate between the collision and subduction segments?Slab break-off at about the transition between continental and oceanic crust in the Alps frees theocean slab in the Carpathians for further subduction. The Carpathian oceanic slab steepens due toits reduced along-strike-length and sets up back-arc extension which is migrating from theAlpine/Carpathian transition E-wards with the back rolling slab. The strike-slip boundary isestablished due to:a) Rollback of the slab along the continental margin establishing a lateral transfer zone

(Fig. 3a & b);b) Lateral tectonic escape of crustal wedges from the now intra-continental indentation site in

the Eastern Alps towards the subduction zone in the Carpathians (Fig. 3b & c);c) Lateral (along strike) flow of gravitationally instable material from the collision site in the

Eastern Alps towards the Carpathians.3. How does the collision to strike-slip transition manifest itself in the upper crust?

Slab break-off in the Alps at ~ 40-35 Ma is manifested by the about coeval closure of isotopicsystems in HP eclogitic oceanic crust and adjacent HP European continental margin basement allalong the strike-of the Alps; this might reflect the onset of exhumation. Slab break-off is furthermanifested by the linear E-W trend of magmatites of ~ 35-25 Ma age along the strike of the Alps;magmatism carries a crust contaminated mantle signature (tonalites). Slab break-off may also beone reason for gravitational orogenic collapse of the thickened and continuously thickening crustunder ongoing intracontinental N-S indentation in the Eastern Alps. Coeval with the collapse theAlpine crust fragments and a lateral tectonic escape fault pattern is set up, along which crustblocks are expelled towards the weakly confined lateral boundary, the Carpathian subduction. Wedefine the interaction of collapse and tectonic escape as lateral extrusion, which thus encompassesplane strain horizontal motion of tectonic wedges driven by forces applied to their boundaries andspreading away from overthickened, rheological weak crust in a orogenic belt towards a weaklyconfined lateral boundary.

4. What is the structural record of the collision to subduction transition?The lateral extrusion process, i.e. the collision - subduction transition, governs the late Oligocene-Miocene tectonic style of the Eastern Alps and Western Carpathians. In the Eastern Alps from Wto E a contractional province passes into a contractional wrench province, which in turn passesinto a wrench-extension province (Fig. 3a & c). Structures related to lateral mass flow from theEastern Alps towards the Carpathian subduction characterize the western part of the Carpathiandomain. The area where the foreland steps towards north, the Western Carpathians, acted as adisplacement transfer zone of the extruding mass; the transfer zone is characterized by a lack ofcrustal thickening. Thickening in the eastern Western Carpathians is localized along en-echelon,right-(N)-stepping thrust segments connected by sinistral strike-slip faults; it is localized wherethe foreland assumes a NW-trend and the extruding mass collided with the foreland.

5. Which boundary conditions are most important in the establishment of the collision tosubduction transition?Experimental models using rock analogues predict the significance of lateral extrusion structuresin the Eastern Alps in terms of gravity spreading (due to thickening in the Alps and back-arcthinning in the Carpathian area) or indentation/escape (due to indentation in the Eastern Alps)(Fig. 3c). They also classify the importance of the boundary conditions which govern thestructural architecture of the Eastern Alps. Ranking from more to less important, these boundaryconditions are: weak lateral confinement in the E (i.e. the transition from a collision site to anorthogonally oriented low-stress subduction boundary), low strength of crust in the Eastern Alps,high strength of the foreland, shape of the indenter in the Eastern Alps, and indentation velocity.

exp1-33

Foreland

unconstrained margin, stiff foreland, straight indenter, 2.5 cm/hBoundary conditions:

Indenter

(a)

(c)

0 100 km

Recent position of foredeep and accretionary wedge

Pre-Neogene rocks

Main fault zone

Active suture zone

Ocean (ocean. crust & transition zone)

(b)

Fig. 3: (a) Deformation structures inthe Eastern Alps and WesternCarpathians during lateral extrusion;(b) Plate-tectonic situation duringearly Miocene (16-17 Ma);(c) Analogue modelling of lateralextrusion.


93

Stress directions in the shallow part of the Hikurangi subduction zone, NewZealand, from the inversion of earthquake first motions

Peter McGinty, Martin Reyners & Russell Robinson

Institute of Geological & Nuclear Sciences, PO Box 30-368, Lower Hutt, New Zealand( [email protected] , [email protected] , [email protected] )

The Hikurangi subduction zone along the east coasts of the North Island and northern SouthIsland of New Zealand is unusual amongst subduction zones worldwide, in that the plateinterface is only about 15 km deep at the coastline. We thus have an opportunity to study theshallow part of the plate boundary, including the seismogenic zone which is capable ofproducing large subduction thrust earthquakes, using land-based observations. In recent yearswe have exploited this opportunity by deploying a number of dense portable seismographnetworks along the subduction zone. The numerous earthquakes recorded during theseseismograph deployments provide an opportunity to investigate stress directions in the shallowpart of the subduction zone. Here we invert for the stress tensor orientation for various parts ofboth the overlying and subducted plates at both ends of the Hikurangi subduction zone - thenorthern South Island and southernmost North Island in the southwest, and the RaukumaraPeninsula in the northeast. To determine stress tensor orientations, we consider all first-motion polarity observations ofselected groups of earthquakes together, regardless of whether or not they are sufficient todefine single event focal mechanisms. It is essentially a grid search over a range of stress axesorientations defined by the greatest principal stress ((1) azimuth and dip, and the least principalstress ((3) azimuth. These parameters are sufficient to define two planes with maximumCoulomb failure stress, and the direction of maximum shear stress on them, assuming acoefficient of friction. For all earthquakes in a selected group, we simply count the number ofpolarities for which observation and prediction are the same, and take the best stress orientationwhich maximizes this count. Error estimates in the stress axes (95% confidence limits) areobtained by the resampling method. The stress inversion suggests which of the two planeswith maximum Coulomb failure stress is the preferred fault plane, as generally more firstmotions will fit one plane than the other. We group hypocentres based on known tectonicboundaries and provinces, while ensuring that there are sufficient polarity observations in eachgrouping to give stable estimates of stress directions. Hypocentres and raypaths have beendetermined using three-dimensional seismic velocity models. Analysis of over 12,500 first motions has provided the following insights into stressdirections in the shallow part of the subduction zone:• In the crust of the subducted plate beneath the southernmost North Island and Raukumara

Peninsula regions, (3 is closely aligned with the dip of the plate. The subducted plate isacting as an efficient stress guide, with slab pull from the deeper part of the plate beingtransmitted to shallow depth.

• Beneath the northern South Island, (3, while still in the plane of the subducted crust, isrotated clockwise out of the downdip direction. This can be related to stronger coupling atthe plate interface in this region.

• Normal faulting on steeply dipping fault planes is favoured in the crust of the subductedplate, suggesting that bending of this crust is accomplished through bulk simple shear (akinto shuffling a vertical deck of cards). Such a bending mechanism has been suggested bystudies of large earthquakes in other subduction zones, and avoids the curling of thesubducted plate inherent in pure bending mechanisms.


94

• In the overlying plate in the northeastern South Island, (1 and (3 are horizontal, thusfavouring strike-slip faulting. The orientation of (1 is close to that expected if the overlyingplate is contracting in response to coupling at the underlying plate interface. However,sinistral motion on NW faults is favoured, whereas motion on the major NE-ENE surfacefaults is predominantly dextral.

• In the Raukumara Peninsula, thrust faulting on steeply dipping planes is favoured in thelower part of the overlying plate and near the plate interface. The stress regime at the plateinterface does not favour interplate thrusting, and a weak interface is required for this tooccur.

• It appears that strain in the overlying plate in both the northeastern South Island andRaukumara Peninsula is partitioned in time - conjugate faults may be active in taking upstrain in the interseismic period between large events on the major faults.


95

The coupled zone of the plate interface at the Hikurangi subduction zone, NewZealand, and its relationship to structure in both the subducted and overlyingplates

Martin Reyners

Institute of Geological & Nuclear Sciences, PO Box 30-368, Lower Hutt, New Zealand ([email protected])

The Hikurangi subduction zone along the east coasts of the North Island and northern SouthIsland of New Zealand is geometrically uniform, and there is little variation in the obliquerelative plate motion along its length. However, there are strong lateral variations in the wayplate motion is accommodated along the subduction zone, with back arc spreading in the northchanging to continental collision 600 km to the south. Much of this change results fromsubduction of the volcanic Hikurangi Plateau on the Pacific plate. The crust of this plateauthickens from about 10 km in the north to about 15 km adjacent to the Chatham Rise in thesouth, and it is thus much more buoyant than the ~7 km thick oceanic crust typically found atother subduction zones. This additional buoyancy has resulted in exposure of the forearc above the shallow part of theplate interface. We thus have an unusual opportunity to study the seismogenic zone of the plateinterface in detail, using land-based observations. In recent years we have exploited thisopportunity by deploying a number of dense portable seismograph networks along thesubduction zone. These have provided insights into the structure and seismic strain regime ofthe subducted and overlying plates, and the nature of plate coupling at the shallow part of theplate interface. At the southern end of the subduction zone in the northern South Island, an earthquakearrival-time inversion for three-dimensional structure images the subducted plate as a relativelylow-velocity feature in the uppermost mantle, reflecting the thick crust which is beingsubducted. Seismic velocity variations within the subducted and overlying plates appear to bespatially correlated. This suggests an interaction between the plates which extends well beyondthe plate interface, and is consistent with other evidence that the plate interface in the northernSouth Island may be permanently locked. We do not see any low-angle thrust earthquakes nearthe plate interface in this region. There is an abrupt decrease in the thickness of the crust (and hence buoyancy) of thesubducted plate northwards across Cook Strait, between the North and South Islands. In thesouthern North Island, we do see low-angle thrust events near the plate interface. Theseconcentrate in two areas: 22-25 km deep beneath Wellington, and about 15 km deep near theeast coast. Through comparison with elastostatic dislocation modelling of GPS measurements,we interpret these thrust events as marking the downdip and updip edges of a locked zone atthe plate interface. The plates appear to be strongly coupled over a downdip width of about 70km, and subduction thrust earthquakes of about MW 8.0 are estimated for this region. By tracking the distribution of small low-angle thrust events northward along the subductionzone, we find that the width of the inferred locked portion of the plate interface progressivelydecreases, to be only about 20 km wide in the northern part of the Raukumara Peninsula.Subduction thrust events of only MW 6.9 are estimated for this region. Also, earthquakearrival-time inversions for 3-D crustal structure indicate a northward increase in the thicknessof sediment carried down by the subducted plate, which in turn can be related to a northwarddecrease in coupling at the plate interface. While changes in coupling in the south of thesubduction zone arise principally from changes in the thickness in the subducted plate, in thenorth they are mainly due to changes in thickness of the overlying plate. When this overlyingplate crust is thin, subducted sediment ponds against relatively strong uppermost mantle, while


96

when it is thicker, the relatively weak lower crust allows sediment subduction to greaterdepths. So we move from a permanently coupled plate interface in the northern South Island,through a 70 km-wide, strongly coupled zone in the southernmost North Island, to a 20-kmwide, weakly coupled zone in the northern part of the Raukumara Peninsula, over a distance ofonly 600 km. This transition is reflected in strike-slip deformation of the overlying plate, withthe intensity of this strike-slip faulting being directly related to plate coupling. In the northernSouth Island, the dextral Marlborough fault system takes up all the margin-parallel platemotion. In the southernmost North Island, the North Island dextral fault belt takes up 75% ofthis motion, and this percentage decreases northwards to be less than 5% in the northern part ofthe subduction zone. The resulting deficit in margin-parallel motion in the north isaccommodated by clockwise rotation associated with back arc spreading. Such spreading isitself promoted by the weak coupling of the plate interface in the north. Because of the rapidlateral changes in plate coupling, the concept of partitioning of oblique relative plate motion isnot particularly applicable to the Hikurangi subduction zone. The eastern strand of the North Island dextral fault belt parallels the downdip edge of thelocked zone determined seismologically. Also, dislocation modelling indicates that the largeearthquakes on this strand in 1855 (M 8.1-8.2), 1931 (MS 7.8) and 1932 (MS 6.9) all occurredon faults which meet the plate interface close to the downdip edge of the inferred locked zone.The geometrical relationship between these faults in the overlying plate and the locked zone isno coincidence. Modelling of Coulomb failure stress changes indicates that thrusting on theplate interface will promote the slip observed on the eastern strand of the North Island dextralfault belt.


97

EAST TO WEST TRANSITION FROM ORTHOGONAL, TO OBLIQUE, TOSTRIKE-SLIP, TO ORTHOGONAL SUBDUCTION ALONG THENORTHERN RIM OF THE PACIFIC BASIN

David W. Scholl

Department of Geophysics, Stanford University, Stanford, CA 94305

From east (Gulf of Alaska) to west (Kamchatka) along the Alaska-Aleutian subduction zone,the relative direction of convergence between the Pacific plate and the leading edge of the NorthAmerica plate changes progressively from orthogonal, to oblique, to strike slip, to orthogonal(Figure 1). The principal plate-boundary processes that effect regional-scale tectonism are,from east to west, (1) continental terrane collision of the Yakutat block in eastern Gulf ofAlaska, (2) disruption of the Aleutian Arc into clockwise rotating and westward translatingblocks of arc massif, (3) arc-parallel splintering of the arc massif, and (4) collision of theseelongated fragments with the Kamchatka subduction zone. Block rotation and length-shearingoccur in response to the westward increase in the right-lateral, plate-boundary shear couple andstrain partitioning. Westward shuttling of arc fragments takes place because they are free tomove in this direction toward the "free surface" of the Kamchatka subduction zone. Arc-continental collision takes place because the NAM-PAC plate boundary has jumped behind(north) of the far western or Komandorsky sector of the arc (Figure 1). In effect, this sectionof the arc is moving toward Kamchatka with much of the speed of the Pacific plate. Thefragmentation style and displacement geometry of block movements imply that the speed ofwestward translation increases in this direction. The critical transition from oblique to strike-slip tectonism takes place along the central andwestern sectors of the Aleutian Arc (Figure 2; Geist et al 1988, 1992 and Geist and Scholl,1994). The process of arc break up and westward translation of blocks and arc splinters hasbeen underway for at least the past 55 my, when the geometry of the present plate boundarysetting was established. However, arc fragmentation and translation appear to have acceleratedduring the past 5-7 my, coincident with the progressing collision of the Yakutat block with thenortheastern margin of the Gulf of Alaska. The modern or late Neogene episode of arcvolcanism along the Aleutian Ridge also got underway at this time. As shown diagrammatically on Figure 2, rotating blocks of arc massif tear away from thearc-magmatic front, leaving trailing-edge basins in their wake. Collision of the Yakutat block,which rapidly elevated and glaciated Alaskan coastal ranges, flushed large volumes of clasticdebris westward along the Aleutian Trench to the front of the arc massif. This resulted in thegrowth of a large accretion wedge and the subduction of thick masses of sediment.Speculatively, the late Cenozoic outbreak of arc volcanisms and accelerated fragmentation andmobilization along the transition zone from oblique to strike-slip plate boundary contact are co-linked to a consequent improved plate-boundary coupling and enhanced injection of subductedfluids into the mantle beneath the magmatic front.

Geist, E. L., Childs, J. R., and Scholl, D. W., 1988, The origin of summit basins of the Aleutian Ridge: implications for block rotation of an arc massif: Tectonics, v. 7, p. 327-341.Geist, E.L., and Scholl, D.W., 1992, Application of continuum modes to deformation of the Aleutian Island Arc: Journal of Geophysical Research, v. 97, p. 4953-4967.Geist, E.L., and Scholl, D.W., 1994, Large-scale deformation related to the collision of the Aleutian Arc with Kamchatka: Tectonics, 13, 538-560


98

0

ALEUTIAN

QUEEN CHARLOTTE F.

RIDGE

KOBUK

KALTAGTINTINA

DE

NA

LI

DENALI

CNT

MUL.

TLKTOT.CHIT.

BEVR

IDIT

.-NX.CH

IRSK

CST

LE. M

T

KAMCHATKAKORYAK

SIBERIA

BERINGIAN

ALEUTIAN

CHAT

km 1000

GULFOF ALASKA

MARGIN

BNB

ALASKA

BR

90KM/MY

YAKUTAT BLOCKCOLLISION

ARC-CONTINENTCOLLISION

ARC SPLINTERING& WESTWARDTRANSLATION

Komandorsky

sector

65KM/MY

FIGURE 1: PLATE TECTONIC SETTING FOR NORTHERN RIM OF PACIFIC BASIN. FROM EAST TO WEST,PROGRESSIVELY FROM ORTHOGONAL TO OBLIQUE TO STRIKE SLIP TO ORTHOGONAL PLATE-BOUNDARY

CONTACT. ARROWS SHOW RELATIVE CONVERGENCE DIRECTION AND RATE

CW BLOCK ROTATION &WESTWARD TRANSLATION

FIGURE 2 : TRANSITION FROM LEFT-OBLIQUE CONVERGENCE CAUSING CW ROTATION AND WESTWARD TRANSLATION OF ARC BLOCK TO RIGHT-LATERAL, ARE-PARALLEL SHEARING AND

WESTWARD TRANSLATION OF ARC CRUSTAL SLICES


99

Tectonic transitions in Papua New Guinea and adjacent seas

Eli A. Silver, Laura Wallace, and Colleen Stevens

Earth Sciences Dept. and Institute of Tectonics, University of California, Santa Cruz, CA 95064

The northern margin of Australia is undergoing complex tectogenesis as a result of arc-continent collision and oblique convergence of the Pacific plate relative to Australia. In theeastern part of the New Britain/Solomon arc system, the Ontong-Java plateau collision hasresulted in reversal of subduction polarity and tearing of the arc, leading to the development ofrapid backarc spreading in the Bismarck Sea (Taylor, 1978). The latter spreads obliquely aswell, in the manner of a leaky transform fault system. Where this transform fault intersects themargin of Papua New Guinea, near the town of Wewak, the tectonics of PNG changesdramatically. To the west, much less detailed study has been carried out, but a new burst ofactivity has accompanied the July 17, 1998 earthquake and tsunami. West of Wewak obliqueconvergence is partitioned into convergence along a narrow trench slope with low sedimentinput, and strike slip faulting onshore (Hutchinson and Norvick, 1978). The July 17earthquake apparently broke along a low angle thrust surface and its slip vector is much morenormal to the coastline than is the plate convergence direction (Harvard and USGS PDE,1998). The large tsunami may have been generated by a large landslide, for which evidence isbeing evaluated at present. Farther west in Irian Jaya, slip also appears partitioned betweenlarge strike-slip earthquakes within the mountain range (Abers and McCaffrey, 1988) andsubduction along the north margin of the range (the north New Guinea trench). Noconvergence is measured with GPS at the south margin of the range (Puntodewo et al., 1994). East of Wewak, thrusting dominates and strike-slip faulting is largely restricted to faultsnormal to the range (Abers, 1989; Abers and McCaffrey, 1994). Transitions occur across thecollision front from oceanic subduction with evidence of subduction erosion in the SolomonSea to rapid accretion, collision and uplift of the mountains as continental crust is overridden(Abbott et al., 1994). GPS studies show rates of convergence of 90-120 km/My across theSolomon Sea, decreasing rapidly eastward into the collision zone (Tregoning et al., 1998).Earthquake locations, GPS observations, and gravity modeling attest to the fact that thick-skinned deformation governs the accretion of the Finisterre Range terrane to the Australiancontinent (Abbott et al., 1994; Stevens et al., 1998; Wallace et al., 1998). During October,1993, a set of large earthquakes occurred at 18-20 km depth and GPS measurements documentthat co-seismic slip occurred very near the ground surface along a steep thrust ramp. GPSmeasurements taken before and after these earthquakes have captured the inter-, co-, and post-seismic slip rates across the frontal thrust (Stevens et al., 1998).

References Cited:

Abbott, L.D., Silver, E.A., and Galewsky, J., 1994, Structural evolution of a modern arc-continent collision in northern Papua New Guinea, Tectonics, 13:1007-1034.Abers, G.A., 1989, Active tectonics and seismicity of New Guinea, Ph.D. thesis,

Mass. Inst. Of Technol., Cambridge, 255 pp.Abers, G.A., and McCaffrey, R., 1988, Active deformation in the New Guinea fold and thrust

belt: Seismologic evidence for strike-slip faulting and basement-involved thrusting, J.Geophys. Res., 93:13,332-13,354.

Abers, G.A., and McCaffrey, R., 1994, Active continent collision: Earthquakes, gravityanomalies, and fault kinematics in the Huon-Finisterre collision zone, Papua New Guinea,Tectonics, 13:227-245.


100

Hutchinson, D.S., and Norvick, M., 1978, Wewak, Papua New Guinea, 1:250,000Geological Series – map and explanatory notes, Geol. Survey of Papua New Guinea, 34pp.

Puntodewo, S.S.O., R. McCaffrey, E. Calais, Y. Bock, J. Rais, C. Subarya, R. Poewariardi,C. Stevens, J. Genrich, Fauzi, P. Zwick, S. Wdowinski, 1994, GPS measurements ofcrustal deformation within the Pacific-Australia plate boundary zone in Irian Jaya,Indonesia, Tectonophysics, 237:141-153.

Stevens, C., McCaffrey, R., Silver, E.A., Sombo, Z., English, P., and van der Kevie, J.,1998, Mid-crustal detachment and ramp faulting in the Markham Valley, Papua NewGuinea, Geology, 26:847-850.

Taylor, B., 1978, Bismarck Sea: Evolution of a back-arc basin, Geology, 7:171-174.Tregoning, P., et al., 1998, Estimation of current plate motions in Papua New Guinea from

Global Positioning System measurements, J. Geophys. Res., 103:12,181-12,203.Wallace, L., Abbott, L.D., Stevens, C., Silver, E., Semer, I., Davies, H., and Cook, T.,

1998, Constraints of subsurface structure of a young arc-continent collision using gravity(abs.), EOS trans. Am. Geophys. Un., 79:F863.


Alpine-Carpathian working groups from the Geological and the Geophysical Institutes atthe Universities of Karlsruhe, Tübingen and Würzburg (Germany)

The Alpine-Carpathian orogenic belt offers an outstanding opportunity to study twosubduction to strike-slip transition zones which both have been active during Tertiary time. Inorder to give a coherent picture of the data and models available for this region (Fig. 1), wedecided to combine our abstracts in one paper with different sections for the differentpresentations. In detail these are: (1) an overview about the Tertiary tectonic evolution of theAlpine-Carpathian orogenic belt, (2) & (3) the kinematics of the two subduction to strike-sliptransition zones, one in the Eastern Alps-Western Carpathians, the other in the Southern andEastern Carpathians, (4) & (5) geophysical data available from the southeastern transition zoneincluding gravity, seismic tomography, recent stress field and seismicity.

Fig. 1: Tectonic overview of the Eastern Alpine-Carpathian region

Bucuresti

49°

48°

47°

46°

45°

44°13° 14° 15°

17° 18°16°19° 20° 21° 22° 23° 24° 25° 26° 27° 28°

44°

45°

46°

47°

48°

49°

50°

51°27° 28°26°25°24°23°22°21°20°19°18°

Budapest

Danube

Tisia

Vienna

ViennaBasin

DanubeBasin

Apusenimountains

TranssylvanianbasinMid-Hungarian fault zone

N

0 100 km

Molasse foredeep

Outer Carpathian flysch nappes

Pieniny Klippen Belt

Pre-Neogene rocks

Neogene calc-alkaline volcanic rocks

Western Carpathians

Eastern Carpathians

Souther

n Carp

athian

s

50°

51°17°16°15°14°13°

18

1615 14 13

12

11.5

9

400 km

North Pannonianblock

Accretionary wedgeEastern Alps

Southern Alps(indenter)

Pannonian

basin

Tisia-Daciablock

8

Age and direction of youngest thrusts onto the foreland

18


Part 1: Lateral extrusion, slab-break-off and subduction retreat: the Oligocene-Recent collision-subduction transition in the Alps and Carpathians

Blanka Sperner1 (now 2), Lothar Ratschbacher3, Peter Zweigel1 (now 4), Franz Moser1, StefanHettel2, Frank P. Lorenz2, Radu Girbacea1 (now 5), Friedemann Wenzel2

1 Geological Institute, University of Tübingen, Sigwartstr. 10, 72076 Tübingen, Germany2 Geophysical Institute, University of Karlsruhe, Hertzstr. 16, 76187 Karlsruhe, Germany3 Geological Institute, University of Würzburg, Pleicherwall 1, 97070 Würzburg, Germany4 IKU Petroleum Research, S. P. Andersens Vei 15 B, 7034 Trondheim, Norway5 Rock Fracture Project, Geol. and Environ. Sciences, Stanford University, CA 94305-2115, USA

New structural, sedimentological, paleomagnetic, and seismological data and analoguematerial modeling outline a geodynamic model for the Neogene tectonic evolution of the Alpine-Carpathian-Pannonian region encompassing two subduction to strike-slip transition zones.Driving mechanisms for the Miocene movements of the Alpine and Carpathian plates were thelateral extrusion in the Eastern Alps and the subduction retreat along the Carpathian arc. Majordriving forces for lateral extrusion in the Eastern Alps are the convergence between the Europeancontinent and the Adriatic microplate which caused the indentation of the Southern Alps andsubduction rollback in the central and eastern Carpathians which exerted a tensional stress on theeastern boundary of the Alps. The particular tectonic architecture of the Eastern Alps is furtherinfluenced by a rigid foreland, a strength minimum in the eastern Alpine lithosphere due tothermal relaxation after the Alpine stacking and a rotating intender.

Subduction retreat started during Oligocene continental collision in the Alps and slab break-off beneath the Alps lead to the reduction in the width of the remaining slab. Resulting sidewaysasthenospheric flow around the edges of the slab reduced the hydrodynamic suction betweenupper and lower plate, so that the slab started to steepen and roll back. Due to the northeasterlyconvex embayment in the orogenic foreland, collision of African continental fragments withEurope migrated from the Alps towards the NE and finally SE (Fig. 2). Those parts of thesubducting slab which were blocked by continental collision detached. The very last stage is welldocumented by seismicity and seismic tomography outlining the youngest slab segment beneaththe SE bend of the Carpathian arc.

In our contributions we review the structures related to lateral mass flow from the Alpstowards the Carpathians, which emphasize a transition from lithospheric thickening to strike-sliptransfer and lithospheric extension in the Eastern Alps (part 2). The Western Carpathians acted asa displacement transfer zone of the extruding mass, which is characterised by a lack of crustalthickening. The Eastern Carpathians are unaffected by the mass extruding from the Alps andshow frontal sediment accretion coeval with strong back-arc extension in the Pannonian realm.Simultaneously, dextral strike-slip movements along the Southern Carpathians accompanied therotation and translation of the Tisia-Dacia-block (Fig. 1 & 2; part 3). The (former) subduction tostrike-slip transition zone at the southeastern bend of the Carpathian arc is identical with the areaof recent subcrustal seismicity. Newest seismic tomography in this region outlines the currentposition of the last remaining slab segment and its ongoing break-off (part 4). Strong seismicityoutlines the dynamics of this process: during the last 60 years four earthquakes with magnitudesbetween 6.9 and 7.7 occurred at depths of 70-180 km. Focal mechanisms, also from smallerearthquakes, prove a strong vertical extension within the subducted slab (part 5).

Plio

cene

- Q

uarte

rnar

y (1

-3 M

a)

51°

50°

49°

48°

47°

46°

45°

44° 13

°14

°15

°17

°18

°16

°19

°20

°21

°22

°23

°24

°25

°26

°27

°28

°44°

45°

46°

47°

48°

49°

50°

51°

27°

28°

26°

25°

24°

23°

22°

21°

20°

19°

18°

17°

16°

15°

14°

13°

010

0 km

late

Ege

rian

(20-

22 M

a)

51°

50°

49°

48°

47°

46°

45°

44° 13

°14

°15

°17

°18

°16

°19

°20

°21

°22

°23

°24

°25

°26

°27

°28

°44°

45°

46°

47°

48°

49°

50°

51°

27°

28°

26°

25°

24°

23°

22°

21°

20°

19°

18°

13°

20°

20°

010

0 km

17°

16°

15°

14°

Pre-

Neo

gene

rock

s

Mai

n fa

ult z

one

Act

ive

sutu

re z

one

Rec

ent p

ositi

on o

f for

edee

p an

d ac

cret

iona

ry w

edge

Oce

an (o

cean

. cru

st

& tr

ansi

tion

zone

)

Ottn

angi

an -

Kar

patia

n (1

6-17

Ma)

51°

50°

49°

48°

47°

46°

45°

44° 13

°14

°15

°17

°18

°16

°19

°20

°21

°22

°23

°24

°25

°26

°27

°28

°44°

45°

46°

47°

48°

49°

50°

51°

27°

28°

26°

25°

24°

23°

22°

21°

20°

19°

18°

17°

16°

15°

14°

13°

40°

10°

010

0 km

Sarm

atia

n (1

2-13

Ma)

51°

50°

49°

48°

47°

46°

45°

44° 13

°14

°15

°17

°18

°16

°19

°20

°21

°22

°23

°24

°25

°26

°27

°28

°44°

45°

46°

47°

48°

49°

50°

51°

27°

28°

26°

25°

24°

23°

22°

21°

20°

19°

18°

17°

16°

15°

14°

13°

010

0 km

Fig.

2: G

eody

nam

ic e

volu

tion

of th

e Ea

ster

n A

lpin

e - C

arpa

thia

n re

gion

(a)

(b)

(c)

(d)


Part 3: The subduction to strike-slip transition in the SE Carpathians: structural,sedimentary, and surface uplift response to subduction retreat and slab break-off

Blanka Sperner1, Franz Moser2, Peter Zweigel3, Radu Girbacea4, Lothar Ratschbacher5

1 Geophysical Institute, University of Karlsruhe, Hertzstr. 16, 76187 Karlsruhe, Germany2 Geological Institute, University of Tübingen, Sigwartstr. 10, 72076 Tübingen, Germany3 IKU Petroleum Research, S. P. Andersens Vei 15 B, 7034 Trondheim, Norway4 Rock Fracture Project, Geol. and Environ. Sciences, Stanford University, CA 94305-2115, USA5 Geological Institute, University of Würzburg, Pleicherwall 1, 97070 Würzburg, Germany

Plate tectonicsDuring Miocene subduction the two intra-Carpathian blocks moved independently with

different directions and velocities, only confined by the geometry of the continental embaymentinto which they moved and which was bordered by the European continent in the north and theeast, and the Moesian platform in the south (Fig. 1 & 2). Rotations of the blocks are proved bypaleomagnetic data which reveal a 30° counterclockwise rotation of the North Pannonian block(NPB) and 60° clockwise rotation of the Tisia-Dacia block (TDB) since the beginning ofMiocene. Continental collision with following slab break-off occurred first in the northernmostpart of the arc and later on shifted towards the SE and S leading to a corresponding shift offoreland basin depocenters and of volcanic activity.

Deformation structures at the frontal part of the TDBA typical accretionary wedge evolved during subduction at the frontal parts of the intra-

Carpathian blocks. Oblique collision of the TDB with the European foreland resulted in differingpost-collisional deformation styles in the northern and southern part of this fold-and-thrust belt.Overthrusting of the northeastern edge of the TDB onto the foreland led to shortening and crustalthickening in this part of the collision zone (Fig. 2c). Simultaneously, thinned continental crustwas still available in the southern part, i.e. in the corner between Moesian platform and Europeancontinent, thus enabling gravitational collapse and basin formation in this part of the accretionarywedge (Fig. 2d). This process was additionally supported by delamination of the lowerlithosphere followed by asthenospheric rise and surface uplift of this area.

Deformation structures at the southern border of the TDBDextral strike-slip movements dominated the Miocene deformation along the contact between

TDB and Moesian platform. Timing of this deformation comes from the subsidence ofsedimentary basins caused by bendings and offsets along the main strike-slip faults (Fig. 4). Lateron, basin inversion and thrusting onto the northern margin of the Moesian platform demonstratethe transpressive character of this strike-slip zone.

Subduction to strike-slip transition zoneForedeep basins evolved around the whole Carpathian arc, but a special situation is given in its

southeastern corner where about 9 km of sediments were deposited during middle Miocene toPleistocene time. Additionally, this region is characterized by strong subcrustal seismicityindicating the presence of a remnant of the subducted slab. Surprisingly, this slab is located80-100 km SE of the Miocene suture zone - an additional indication for delamination androllback of the lower lithosphere (Fig. 5).

Recent surface upliftLarge uplift rates are recorded from the northern part of the Eastern Carpathians where the

TDB overthrusted the European foreland and thus caused crustal thickening. Smaller, but stillremarkable uplift rates occur in the southern part where no crustal thickening took place, so thatadditional processes have to be considered. Slab break-off which started in the N and migratedtowards the S can serve as such an additional mechanism.

Fig. 4: Overthrusting of the Tisia-Dacia block onto the European foreland in the NE and onto the Moesianplatform in the S resulting in crustal thickening. Timing of basin subsidence, followed by inversion andoverthrusting.

Fig. 5: Profile through the Eastern Carpathians (s. Fig. 4) showing delamination of the lower lithopshere

ESEWNW Foredeep

40

60

80

100

120

140

160

180

20

km

Slab pull

Asthenospheric rise and decompression

AsthenosphäreLithosphärischer MantelOzeanic crustContinental crust

Alkaline magmatism

Calc-alkaline magmatism

Miocene suture

Tisia-Dacia block European plate

Earthquake

Volcanism

Accretionary wedge

50 km

N

0 100 km

2

1-4

Tisia-Dacia block Thickened crust

Getic depression3

1-4

Basin subsidence during 1 - late Egerian, 2 - Ottnangian, 3 - Sarmatian, 4 - Pliocene

Basin inversion and / or thrusting during 1 - late Egerian, 2 - Ottnangian, 3 - Sarmatian, 4 - Pliocene

1-2 2-33

4

Profile in Fig. 5

former continental margin

"Normal" crust{


106

Plate deformation at depth under northern California

Uri S. ten Brink1, Nobumichi Shimizu2 and Philipp C. Molzer3

1US Geological Survey, Woods Hole, MA 02543 USA. Tel. 508-457-2396. Fax: 508-457-2310. E-mail:[email protected] Hole Oceanographic Institution, Woods Hole, MA 02543 USA3US Geological Survey, Woods Hole, MA 02543 USA

The prevailing view of asthenospheric upwelling (a slab gap) undercoastal California stems from the assumption that plate boundaries atdepth behave rigidly and are of zero width. For northern Californiaplate kinematics predict a gap in the underlying subducted slab causedby the northward migration of the unstable Pacific-North America-Juan deFuca Triple Junction. Large scale decompression melting andasthenospheric upwelling to the base of the former forearc are notsupported however, by geophysical and geochemical observations. Wesuggest that perhaps a more realistic view of the interaction betweenthe Pacific, Juan de Fuca and North American plates is one where theJuan de Fuca and Pacific plates under coastal northern California deformcontinuously to fill the geometrical gap in the wake of the northwardmigrating Mendocino Triple Junction. We explain the minor Neogenevolcanic activity, its chemical composition and the observed heat flowby thermal re-equilibration of a 20-30 km thick forearc underlain by athinned young Juan de Fuca slab. This view is consistent with the regionbeing underlain by a fossil oceanic crust and by an upper mantlelithosphere of young thermal age. It is also consistent with thepervasive deformation of the southern Juan de Fuca plate and theMendocino Triple Junction region onshore. In contrast, the consequencesof a “true” slab gap in the forearc region can be examined in the InnerCalifornia Borderland offshore Los Angeles where local asthenosphericupwelling took place during the Miocene as a result of horizontalextension and rotation and is manifested by high heat flow, voluminousmagmatism, primitive melt composition, and the absence of a fossilcrust.


107

Liquefaction Features: A Potentially Powerful Paleoseismic Tool in theDominican Republic

Martitia Tuttle1, Carol Prentice2, Luis Pena3, and Kathleen Dyer-Williams4

1Department of Geology, University of Maryland, College Park, Maryland, USA 20742, [email protected]. S. Geological Survey, 345 Middlefield Rd MS 977, Menlo Park, California, USA 94025,[email protected] Cuesta Colorado Esquina Calle No. 9, Las Colinas, Santiago, Dominican Republic4Department of Geology, University of Maryland, College Park, Maryland, USA 20742, [email protected]

Earthquake-induced liquefaction features, including sand blows and sand dikes, occur inthe eastern, central, and western parts of the structurally controlled Cibao valley in northernDominican Republic. This region is traversed by the Septentrional fault, an onshore portion ofthe North American-Caribbean plate boundary. Our observations of liquefaction features inthis region are significant because paleoliquefaction features can be used to estimate the timing,source areas, and magnitudes of prehistoric earthquakes and are particularly useful in areaswhere active faults cannot be studied directly.

Accounts of historic earthquakes in 1842, 1897, and 1946 described ground failure typicalof liquefaction. Scherer (1912) reported that springs and ground cracks formed along the RioYaque in the vicinity of Santiago (central Cibao valley) during the 1842 earthquake. He alsoreported that cracks formed and the ground subsided near Guayubin (western Cibao valley)during the 1897 earthquake. Following the 1946 earthquakes, Lynch and Bodle (1948)reported that high silt banks along the Rio Yuna near Arenoso (eastern Cibao valley) settled interraces paralleling the river and that sand and water spouts occurred up to 100 m from the riverbank. Because liquefaction typically recurs where susceptible deposits are present, sites ofhistoric liquefaction are prime targets for paleoliquefaction studies.

In the central Cibao valley, Prentice et al. (1993, 1994) found liquefaction features in atrench across the Septentrional fault, where a sand dike apparently intruded along a fault plane.Radiocarbon dating suggests that these features formed during rupture of the Septentrional faultcirca A.D. 1200. In the eastern and western Cibao valley, where the active trace of theSeptentrional fault is buried by Holocene alluvial deposits, we conducted reconnaissance forliquefaction features along 13 km of the Rio Yuna near Arenoso and along 20 km of the RioYaque between Guayubin and Castanuela. We selected these areas for reconnaissance basedon the likely presence of liquefiable sediments and availability of cutbank exposure ofHolocene deposits. We documented liquefaction features at seven sites along the Rio Yuna andat ten sites along the Rio Yaque. Multiple features occur at almost every site and we foundevidence of recurrent liquefaction at five of the sites.

At a site along the Rio Yaque near Castanuela in the western Cibao valley, we documentedat least two, and possibly three, generations of liquefaction features, including sand blows andrelated sand dikes. Sand blows range up to 6 cm in thickness and most sand dikes range from1 to 18 cm in width. A 42-cm-wide compound sand dike apparently was utilized by ventingsand and water during more than one event. The youngest generation of liquefaction featuresoccurs high in the section and may have formed during the 1897 earthquake. Unfortunately,we found no material suitable for radiocarbon dating at this site. Sand dikes at other sites alongthe Rio Yaque range from 2 to 15 cm in width. Sand blows may occur at several of these sitesbut additional excavation is needed to verify this. At one of these liquefaction sites, charcoalwas collected near the base of the cutbank from the host deposit. It yielded a calibrated age of6490 to 5930 years B.P. and provides a maximum age of liquefaction features observed alongthe Rio Yaque.


108

Near Arenoso, along the Rio Yuna in the eastern Cibao valley, we logged prehistoric sandblows and related sand dikes at three sites. The sand blows range from 30 to 105 cm inthickness and the sand dikes range up to 24 cm in width. The two thicker sand blows arecompound structures and probably formed during two or more events in an earthquakesequence. Sand blows at the three sites have similar stratigraphic positions to each other andmay have formed during the same earthquake sequence. Radiocarbon dating of charcoal withinone of the sand blows and of soil clasts within two feeder dikes yielded similar results andindicate that the sand blows formed after A.D. 690. Radiocarbon dating of clayey soilhorizons overlying the sand blows yielded ambiguous older dates. Burr (this volume)attributes these older dates to carbonates held by clay particles within the samples. Althoughnot conclusive, radiocarbon dating of these sand blows allows for the interpretation that theyformed during the A.D. 1200 rupture of the Septentrional fault in the central Cibao valley. Incomparison, relatively young sand dikes that may have formed during the 1946 earthquakes aresmaller ( < 7 cm) than prehistoric sand dikes in the area. Furthermore, historic sand blowsappear to be less prevalent than the prehistoric sand blows. These observations suggest thatthe prehistoric earthquakes were larger than or located closer to Arenoso than the historicearthquakes.

Additional study is needed to map the age and size distribution of liquefaction features inthe eastern and western Cibao valley. Age estimates of paleoliquefaction features must benarrowly constrained in order to differentiate earthquake sequences and to correlate featuresacross a region (Tuttle et al., 1996). Regional correlations of liquefaction features areimportant because they establish the area affected by high levels of strong ground shaking,which can, in turn, help constrain models of source areas. In general, the largest features in aregional distribution of paleoliquefaction features reflect the source area of a prehistoricearthquake. A better understanding of the age and size distribution of liquefaction features inthe Cibao valley would help to define the source area and magnitude of the A.D. 1200 event, toidentify other prehistoric earthquakes that struck the region, and possibly to differentiateearthquakes produced by thrusting from strike-slip events.

Liquefaction features provide a tool to help reconstruct the prehistoric earthquake record ofthe Dominican Republic, and thus to assess the earthquake hazard of this seismically activeregion. In addition, our observations indicate that liquefaction and lateral spreading is likely tooccur during future large earthquakes. Defining areas underlain by liquefiable sediment wouldbe an important step towards mitigating this hazard.

Lynch, J. J., and Bodle, R. R., 1948, The Dominican earthquake of August, 1946: Bull.Seismol. Soc. Am., v. 38, 1-17.

Prentice, C. S., Mann, P., Taylor, F. W., Burr, G., and Valastro, S., Jr., 1993,Paleoseismicity of the North American-Caribbean plate boundary (Septentrional fault),Dominican Republic: Geology, v. 21, 49-52.

Prentice, C. S., Mann, P., Burr, G., and Pena, L. R., 1994, Timing and Size of the MostRecent Earthquake Along the Central Septentrional Fault, Dominican Republic [abs.]: in,Prentice, et al., 1994, Proceedings of the Workshop on Paleoseismology: USGS Open-File Report 94-568, p. 158.

Scherer, J., 1912, Great earthquakes in the island of Haiti: Bull. Seismol. Soci. Am., v. 2,161-180.

Tuttle, M. P., and Schweig, E. S., 1996, Recognizing and dating prehistoric liquefactionfeatures: Lessons learned in the New Madrid seismic zone, central United States: Journ.Geophys. Res., v. 101, n. B3, p. 6171-6178.


Tectonic Activity in the Mona Passage Area during the three Cretaceous-Holocenetectonic phases of the Puerto Rico-Virgin Islands platform.

Jean-Paul van Gestel1, Paul Mann2, James F. Dolan3, and Nancy R. Grindlay4

1Department of Geological Sciences, University of Texas at Austin, Austin, TX 78712,[email protected];2Institute for Geophysics, University of Texas at Austin, 4412 Spicewood Springs Road, Bldg. 600,Austin, TX 78759-8500, [email protected];3Department of Earth Sciences, University of Southern California, Los Angeles, CA 90089-0740,[email protected],4Department of Earth Sciences, University of North Carolina at Wilmington, Wilmington, NC 28403-3297, [email protected]

Introduction: In the summer of 1996 single-channel seismic reflection, sidescan, bathymetry, gravity, and magneticdata were collected during cruise EW 96-05 on board of R/V Maurice Ewing in which all authorsparticipated. These newly acquired data have been integrated with and correlated to onland outcrops andwell information and have been used to re-examine the proposed regional geologic and tectonic models.The depositional history of the different sequences of the platform North of Puerto Rico has been studiedand three different tectonic phases have been distinguished.Setting: Puerto Rico occupies the northeastern segment of a lower Cretaceous to Holocene island arc chain thatextends from western Cuba to the north coast of South America. The core of the island is formed byCretaceous arc basement rocks. Both in the north and in the south Oligocene-Early Pliocene carbonateplatform sediments can be found. In the north the platform covers a small portion of the island but extentsunder a smooth continuous 4 degree dip offshore to a depth of 4 km deep. In the south the platform has asteeper dip and it extends less far offshore. The whole shallow submarine area in between the islands ofHispaniola, Puerto Rico and the Virgin Islands is also covered with Oligocene-Pliocene carbonate rocks.The Puerto Rico-Virgin Islands platform overlies the Puerto Rico-Virgin Islands microplate that has beendefined on the basis of earthquake data and marine geophysical sidescan surveys of offshore areas. To thenorth, the microplate is bounded by the Puerto Rico trench. To the east of Puerto Rico, the platform trendsto the east-northeast and covers most of the area of the Virgin Islands. The eastern and southeastern edgeof the platform and the Puerto Rico-Virgin Islands microplate are sharply bounded by the Anegada faultzone. This fault zone runs trough the deepwater Anegada Passage between Puerto Rico and St. Croix andconnects the Sombrero and Virgin Islands basins. These two fault-bounded deeps are pull-apart basinsalong the right-lateral Anegada fault zone. The southern edge of the Puerto Rico-Virgin Islandsmicroplate is defined by the Muertos Trough where the Caribbean plate is subducted beneath themicroplate. To the west of Puerto Rico, the carbonate platform extends across the Mona Passage and ontothe island of Hispaniola. Bathymetric deeps in the Mona Passage correspond to rifts that locally extendand fragment the platform between Puerto Rico and Hispaniola. The origin and age of this rifting in theMona Passage area is unclear and will be the focus of this presentation.Three tectonic phases of the northern margin of Puerto Rico: In a regional study of the offshore seismic reflection profiles surrounding Puerto Rico, we havedocumented three distinct, Cenozoic tectonic phases that occurred during development of the northernmargin of Puerto Rico.Tectonic phase 1: Late Cretaceous to Middle Eocene formation and sedimentary infilling from the southof a forearc basin. This basin was formed between down-to-the-north normal faults near the present daycoast of Puerto Rico and an outer-arc ridge near the present day shelf break. Tectonic phase 1 containedthe last volcanic arc activity in Puerto Rico that was produced by the subduction of oceanic crust of theNorth America plate beneath the Caribbean arc system. The end of tectonic phase 1 is related to an initialcollision between the Caribbean arc and the Bahama carbonate platform.


Tectonic phase 2: Latest Eocene to early Pliocene formation of a 1578 m thick, northward-thickeningPuerto Rico-Virgin Islands platform, predominately formed of carbonate sediments. Depositionalthicknesses of sedimentary layers deposited during phase 2 are controlled by two large arches: 1) TheNNW-trending Guajataca arch appears to have formed as the elevated flank of the north-south strikingMona rift, which forms the western boundary of the present day Puerto Rico-Virgin Islands microplate, 2)the northeast-trending San Juan arch can not be related to any adjacent structure or plate boundary feature.The results of paleomagnetic studies of the carbonate sediments show that Puerto Rico experienced acounterclockwise rotation during tectonic phase 2.Tectonic phase 3: Early Pliocene to Holocene northward tilting of the Puerto Rico-Virgin Islandsplatform, submerged the northern edge of the platform to a depth of 4 km and elevated the southern edgeof the platform to several hundred meters above sea level on Puerto Rico. Northward tilting of this areaoccurred on the northern limb of a large arch or anticline formed parallel to the long axis of the island ofPuerto Rico and its shelf areas. The arch probably formed in response to a post early Plioceneconvergence between the North America and Caribbean plates.Reconstruction measurements in the Mona Passage: Three lines with seismic reflection data have been collected in the Mona Passage area during cruiseEW96-05. Two longer NE-SW lines that cross the Mona and Yuma Rift areas and one east-west line thatonly crosses the Mona Rift. These three lines have been interpreted and reduced to their original lengthusing reconstruction software packages in an effort to calculate their extension. For the East-Westtrending line 31 the extension has been estimated to be 6.35 km. In the two other lines crossing the MonaRift, being line 32 and 35, the estimate is only 4.5 and 3.1 km. However one has to take into considerationthat the orientation of these two lines is not parallel to the estimated main direction of extension whichwill cause the calculated amount to be smaller than the real extension. One has to take into considerationthat these numbers contain a fairly large interpretation uncertainty. Extension of 6.35 km between the twoislands of Hispaniola and Puerto Rico combined with the 2 mm/yr of difference in the eastward movementof the two islands observed in the GPS measurements would lead to a 3.2 million year old Mona andYuma Rift extension.Comparisons between the carbonate platform on both sides of the Mona Rift: In our study of the well, seismic reflection and outcrop data there was a considerable thickness differencenoted in the two sequences of the carbonate platform on both sides of the Mona Rift. In the north coastbasin the maximum thickness of the platform is 1578 m. However towards the Mona Rift it appears thatthe platform is thinning to a minimum of only 600 m on the east-slope of the Mona Rift. When studyingthe isochron maps of the different sequences within the platform this thinning can be found to appear inthe lower and older sequences, while the upper and youngest sections of the platform have a more uniformdistribution. The thinning of the lower sequences onto the edge of the Mona Rift can be explained byactivity of the NNW-trending Guajataca Arch. This arch appears to have formed as the elevated flank ofthe north-south striking Mona rift, which forms the western boundary of the present day Puerto Rico-Virgin Islands microplate.Conclusions: In conclusion, we believe that the Mona Rift has evolved in two different periods. The rifting was firstactive in the second tectonic phase, when the area in between Hispaniola and Puerto Rico functioned asthe pivoting point of the counterclockwise rotation of Puerto Rico. Compared to the homogeneous,constant thickness platform in the north coast basin of Puerto Rico the isochron map of the carbonateplatform and the seismic reflection profiles show a large amount of variation in the Mona Passage area.This is the result of seismic activity during the formation of the carbonate platform. After the formation ofthe platform in tectonic phase 1, the rotation stopped and the arching phase began. During this thirdtectonic phase, the Mona Rift started opening, which seems to agree well with the calculated age of theorigin of the extension.


111

TECTONICS OF THE ALBORAN SEA BASIN: A TRANSTENSIONALBASIN IN A TRANSPRESSIONAL REGIME

J.T. Vázquez (1) and R. Vegas (2)

(1) Dpto. Geología, Fac. Ciencias del Mar, Univ. Cádiz, Campus Rio San Pedro, 11510-Puerto Real. [email protected](2) Dpto. Geodinámica. Fac. Ciencias Geológicas. Univ. Complutense. 28040 - Madrid. Spain.

The Alboran Sea Basin is located at the western edge of the Mediterranean Sea between theIberian Peninsula and northern Africa. It extends westwards of the South Balearic Basin and itis surrounded by the Betic-Rif Arc which represents a collisional orogen and it is included inthe westernmost chains of the Mediterranean Alpine belt. The basin has been originated duringthe Miocene and its formation and later evolution are closely related to the Betic-Rif orogendevelopment and to the oceanic spreading of the Western Mediterranean Basin. A set of structures associated with different stages are superimposed. The main structuraltrends, indicated by the bathymetry, basement structure and volcanic buildings display thefollowing directions ENE-WSW to NE-SW and WNW-ESE. The oldest structures are relatedto normal faults that have been active until the Lower Tortonian. The structures that affect theUpper Miocene to Quaternary units have been interpreted as oblique strike-slip faults. Thedevelopment of some transpressional (ramp anticlines and ridges) and transtensional (pull-apartbasin and local collapses) features which determine the physiography have been identified. The aeromagnetic studies and drills show a continental basement. The gravimmetric, heatflow, refraction and deep reflection data show a thinned crust, about 17 km, which thicknessdecreases eastwards of the Alboran island meanwhile southwards the crust reaches to 23 kmwhich is a consequence of the African margin wide in the eastern part of the basin. A detachedslab below the Betic-Alboran region, between 200-670 km of depth, characterise the mantlestructure. The seismicity is characterised by a continuous activity of moderate to lowmagnitude. It is remarkable the occurrence of intermediate depth events (45-150 km) that drawa narrow arc in the western part of the basin. Several aspects should be pointed out in the tectonic setting of the Alboran Sea Basin:1) It is located on the plate boundary between Eurasia and Africa and its Cenozoic evolution iscontrolled mainly by a convergent regime from the Late Cretaceous to Tortonian times and atranspressive regime from that time to present (Dewey et al., 1989). In this context, acontinental lithospheric block (Alboran Domain) would have collided, simultaneously andlaterally, against the two largest plates in the Mediterranean area, Eurasia and Africa, in thelocation of the main plate boundary. It can not be rejected that previously to the currentcollision a limited subduction of oceanic lithosphere would have developed.2) It overlies a thinned continental crust segment that mainly belongs to the Alboran crustaldomain (Vegas et al., 1995)3) Nowadays the Alboran Sea Basin is an extensional basin located in a backarc position inrelation to the Betic-Rif Arc. However, it does not display all the backarc basins characteristics.It does not exist a B-subduction, either exists volcanic processes related with this subduction,neither oceanic crust floored this basin. There are some calcoalkaline volcanic buildings ofMiocene age, mainly in the eastern sector, but this volcanism is not related to the configurationof a volcanic arc system associated with an oceanic subduction zone.4) It is situated in the western side of an oceanic basin (Western Mediterranean Basin), whoseformation has been related with the extensional collapse of a Paleogene Alpine orogen. TheAlboran Domain was probably structured in this previous orogenic stage and then wasextensionally thinned.


112

There are several proposed models to explain the origin of the Alboran Sea Basin (seediscussion in Docherty and Banda, 1995 and in Lonergan and White, 1997) including basinformation in relation to 1) the western migration of a rigid microplate, 2) escape tectonicprocesses, 3) transtensional origin, 4) backarc basin related to E directed subduction, 5)backarc basin related to a retreating W directed subduction and/or slab pull associated with adetached slab, 6) orogenic extensional collapse related to asthenospheric diapir, crustaldelamination and/or convective removilization. We propose a mechanism of progressive backarc transtension to explain the generation of theAlboran Sea Basin. This process would be related to the Lower to Middle Miocene roll backretreating (towards the east, south and west) of a previous subduction zone NW directed,which corresponds to the splitting by means of gravitational collapse of a Paleogene Alpineorogen, in the context of a N-S convergence between Eurasian and African plates (LateOligocene-Lower Miocene). Two Mediterranean zones have been extruded by lateral escape tectonics to accommodate thisshortening, the Betic-Rif and Calabrian arcs, controlled by embayements (free space) andpromontories influence. In the Alboran region, the upper plate (Alboran Domain) of thissubduction zone migrated westward (about 100-300 km) over an oceanic or very thinnedcontinental lithosphere and collided to Southiberian and Northafrican margin in an obliqueway. This westwards escape tectonic took place in the Lower Miocene and constituted asecondary convergent/transpressional component that generated a collisional arc in the upperplate thrust front, two transpressional zones one in each lateral borders and an extensionalbackarc basin. In this sense the Alboran Sea Basin is generated in the internal part of awestward moving plate coeval with the development of compressional structures in other partsof the region. When the upper plate collide against the normal continental crust of Iberia andAfrica the extrusion mechanism is blocked. The final configuration of the Alboran Sea Basin took place at Tortonian times, when theEurasian-African plate NW-SE convergence is clearly transpressive respect to the possible plateboundary and caused transtensional and transpressional reactivations simultaneously. Also,certain slab pull process could exist in the Western Alboran Basin related to a subducted slablocated around 45-150 km of depth as it show by a Pliocene collapse of the margin.Nowadays, the Eurasian-African transpressional convergence produces internal deformation inthis area and compressional structures in the Gulf of Cadiz.AcknowledgementsThis work has been supported by a CICYT project, AMB 95-1151-E, of the Spanish ResearchProgrammeReferencesDewey, J.F., Helman, M.L., Turco, E., Hutton, D.H.W. & Knott, S.D. (1989) Kinematics

of the western Mediterranean. In: Coward, M.P., Dietrich, D. & Park, R.G. (eds.)Alpine Tectonics. Geological Society Special Publications, 45: 265-283.

Docherty, C. & Banda, E. (1995) Evidence for the eastward migration of the AlboranSea based on regional subsidence analysis: A case for basin formation by delamination ofthe subcrustal lithosphere? Tectonics, 14 (4): 804-818.

Lonergan, L. & White, N. (1997) Origin of the Betic-Rif mountain belt. Tectonics, 16 (3):504-522.

Vegas, R., Vázquez, J.T., Medialdea, T. & Suriñach, E. (1995) Seismic and tectonicinterpretation of ESCI-Béticas and ESCI-Alborán deep seismic reflection profiles: structureof the crust and geodynamic implications. Rev. Soc. Geol. España, 8 (4): 449-460


113

SUBDUCTION AND STRIKE-SLIP (TRANSPRESSION) IN THE BORDERSOF THE ALBORAN, CARIBBEAN AND SCOTIA PLATES: ACOMPARATIVE STUDY

R. Vegas (1), J.T. Vázquez (2) and J. Acosta (3)

(1) Dpto. Geodinámica. Fac. Ciencias Geológicas. Univ. Complutense. 28040 - Madrid. [email protected](2) Dpt. Geología, Fac. Ciencias del Mar, Univ. Cádiz, Campus Rio San Pedro, 11510-Puerto Real. Spain(3) Instituto Español de Oceanografía, C/ Corazón de María, nº 8. 28002 - Madrid. Spain.

At present none of the large lithospheric plates is bounded by two subduction-related orcollision zones with opposite polarity. This fact indicates the predominance of the ridge-push inthe motion of plates. Nevertheless, the small Caribbean and Scotia plates, as well as the nowinactive Alboran platelet, contain two compressive (transpressive) boundaries with oppositevergence joined by a frontal subduction arc. All the three small plates present clear evidence oflarge displacements along the parallel or almost parallel transpressive boundaries. Since thesedisplacements are not related to a centre of expansion, the driven mechanism for these smallplates remain unexplained. The other problem related to this particular plate configurationcorresponds to the area of junction between the frontal subduction of the leading arc and thetranspressive plate-boundary zone. Taking into account that the level of understanding for thethree plates is unequal, it should be useful to compare their common features in order toadvance some hypothesis concerning the problems mentioned previously and some of theirconsequences, such seismicity and internal deformation. The Alboran, Caribbean and Scotia plates occupy the site of former seaways created duringthe first stages of the rupture of the Pangea. Their emplacement astride the main plate-boundaries, including large transport of crustal units and advance at the leading arc, has beenrelated to the convergence of the large plates initiated during the Upper Cretaceous platerearrangement. This sort of extrusion implies strong convergence that is only recorded in theAfrica-Europe trajectory. But even in the case of the Alboran plate, the direction of the Africa-Europe slip vector coeval with the time of the westward displacement seems to be inadequate.Therefore, an ill-understood mantle flow as driving mechanism for this peculiar tectonicbehaviour should be invoked. The other main problem related to these small, intermediate, plates is the nature of theirtranscurrent lateral borders and the mode in which they merge with the frontal subducting arc.This is also a consequence of the origin of the escape of the intermediate plates, prevailingtranscurrent or compressive tectonic regimes. In the Alboran plate the transcurrent borderscorresponds to a crustal overthrust onto the African and Iberian margins and constitute theBetic and Rif ranges. Their present position is obscured by late extensional episodes, althougha short segment of the Africa-Alboran limit has the shape of a strike-slip zone connecting theRif and Tell chains of Northern Africa. Kinematic indicators show clear transcurrent motionalong these margins inducing systematic block rotations along the disrupted cover of theAfrican and Iberian margins. Deformation is at present accommodated by internal deformationof the intermediate Alboran plate, as shown by the distribution of the seismicity. A remnant ofthe frontal seismicity due to the advance of the Gibraltar arc can be found in a N-S zone ofintermediate-depth seismicity that defines a near vertical plane delineating the chord of the arc.However, the main shocks are located outside of the arc, where the oceanic lithospheres of thebig plates meet directly accommodating the convergence in a zone of important submarineuplifts and related basins.


114

Concerning the active, moving intermediate plates Caribbean and Scotia, the transcurrentborders show a remarkable resemblance in the segments where the oceanic lithosphere of thisbig plates is in contact with the overriding crustal slivers of the intermediate plates. Seismicimages of the Puerto Rico Trench and the Scotia-Antarctica Plate Boundary Zone, far from theSouth Sandwich frontal arc, are remarkably similar, showing a limited subduction of the NorthAmerican and Antarctic oceanic lithospheres respectively. This type of non conventionalsubduction implies the absence of volcanism and some sort of strain partitioning in order tomaintain the shallow and limited underthrusting of the oceanic lithosphere. Moreover, in thejuncture of this limited, lateral subduction and the frontal subduction, a sharp change in the dipangle for the lower plate should be considered. This may be supported by the attitude of theremnant seismicity in the inner part of the Arc of Gibraltar. The limited subduction can beconnected with the collision found in the segments of the transcurrent borders where the lowerplate, i.e. the large bordering plate, is formed by continental lithosphere. In this case theoverthrusts resemble the crustal stacking in the Alboran transcurrent borders. This implies thatif the coupling between the two plates is of sufficient extent, the occurrence of big shocks canbe expected. As in the Alboran plate, some segments of the transcurrent borders correspond tozones of predominant strike-slip faults with minor, if any, sign of convergence. Theoccurrence of big shocks in long strike-slip faults can be expected, in particular in theintracontinental parts of these fault corridors. The internal deformation of the intermediateplates, mainly in direct relation with the converging transpressive borders, where crustal sliverscan act as minor subplated, can obscure the kinematic indicators, and hence making impossiblethe determination of the direction of the relative plate motion between the intermediate plate andthe surrounding ones. This constitutes another important problem for the definition of the maincauses of the formation of these intermediate small plates. In any case, these intermediate plateshave many features in common whose origin must be attributed to a characteristic plate setting.The comparison between the few examples, past and present, must help understand theirformation and evolution.AcknowledgementsThis work has been supported by a CICYT project, AMB 95-1151-E, of the Spanish ResearchProgramme.


115

USGS-MIDAS SEISMICITY MAP OF THE CARIBBEAN AND JOINTEFFORTS FOR SEISMOLOGIC COOPERATION

Christa von Hillebrandt-Andrade 1, Carlos Mendoza and Madeleine Zirbes2 and Luis OdonelGomez 3

1. Red Sismica de Puerto Rico, Depto. de Geologia, Universidad de Puerto Rico, PO Box 9017, Mayaguez,PR 00681-9017 ([email protected]) 2. National Earthquake and Information Center, U. S. Geological Survey, Box 25046, M.S. 967, DenverFederal Center, Denver, Colorado 80225 ([email protected] and [email protected]) 3. Instituto Nacional de Recursos Hidraulicos, Santo Domingo, Rep. Dominicana ([email protected])

In 1998 the United States Geological Survey (USGS) and the Middle America Seismographic Consortium (MIDAS) published a seismicity map of the Caribbean region. The area covered by the map extends from 45°to 125° W and 5° S to 35° N. The seismicity data cover the years 1900-1994 and were obtained from the revised historical and instrumental earthquake catalogue with uniform moment magnitudes compiled for the Caribbean and Latin America by the Instituto Panamericano de Geografia e Historia (IPGH). To minimize incompleteness, only earthquakes of magnitude 4.2 or larger are plotted. Earthquakes are classified into three magnitude ranges and four depth ranges. The map shows a total of about 17,000 earthquakes for the time period considered. Active volcanoes of the region are also included as well as all cities with populations over 25,000. Bathymetric features are also shown and highlight important correlations between seismic activity and known tectonic features of the Caribbean. Apparent increases in seismicity can also be seen at plate-boundary transition zones where plate interaction changes from subduction to strike-slip motion (e.g., the Puerto Rico-Dominican Republic area and the Trinidad and Tobago- Venezuela region). Publication of the Caribbean seismicity map is one of several initiatives pursued by MIDAS since its inception in 1990. Other activities include the installation of broadband stations in the Caribbean region, the establishment of an Electronic Seismic Data and Information Center in the facilities of the Puerto Rico Seismic Network in the Department of Geology of the University of Puerto Rico at Mayaguez (PRSN) and the planning of technical workshops for 1999 in Panama and Puerto Rico on broadband data processing and seismic data exchange. These activities have aided MIDAS to achieve its principal objective of promoting scientific interaction among participating institutions and facilitating the rapid, efficient, and widespread utilization of earthquake data by Middle America and Caribbean countries in routine observatory operations and in more advanced seismological research. In 1993, a broadband station (UPIG) was installed in Panama City as a joint effort between the USGS, the Universidad Autonoma de Mexico (UNAM) and the Institute of Geosciences of the University of Panama. More recently in 1997, joint cooperative efforts between the USGS and UNAM led to the installation of a broadband station (LPIG) in La Paz, Baja California, Mexico. In 1998, this station was linked via Internet to the USGS National Earthquake Information Center. Since 1997, the Earthquake Unit of the University of the West Indies, Jamaica and the PRSN have been working towards the installation of a joint broadband station in Jamaica. Through its Home Page (http://midas.uprm.edu), the MIDAS Electronic Seismic Data and Information Center provides interested parties information on the MIDAS Consortium, seismic stations and IPGH and MIDAS catalogues of events located in


116

the MIDAS region (55° to 120° W and 5° S to 33° N). The MIDAS catalogue, which has a search engine, is automatically updated with the hypocentral information provided by the participating institutions. A special section in the webpage additionally provides information about significant earthquakes (magnitude or MM Intensity greater than six) that have occurred in the MIDAS region. For these events, each of the participating institutions contributes phase data and/or broadband waveforms; available moment tensor mechanisms computed for the earthquakes are also placed online as well as the epicentral map and comments and images of the earthquake and its effects. Links to useful earthquake information, data, services, utilities and programs are also included. Funds for the establishment of the MIDAS data center and webserver were secured through the cooperation of various participating institutes. Sources for additional funding are currently being sought to guarantee the maintenance and continual development of the data center.


117

Tectonic origin of "suprastructure-infrastructure" architecture, NorthernRange, Trinidad

Weber, J.1, with contributions by Ferrill, D.2 and Teyssier, C.3

1Grand Valley State University, Allendale, MI 49401 USA

2Center for Nuclear Waste Regulatory Analyses, Southwest Research Institute, San Antonio, TX 78238-5166USA

3University of Minnesota, Minneapolis, MN, 55455 USA

Weber et al. (in review) used temperature-sensitive quartz and calcite microstructures andfission track data to identify, describe, quantify, and map peak deformation temperatures in thelow grade metasedimentary rocks across the Northern Range, Trinidad. Before this study,internal differences in thermal grade across the range had not been resolved (e.g., Frey et al.1988). Deformation temperatures are relatively low in the eastern Northern Range and higherin the central and western Northern Range. Accompanying this east to west increase in thermalgrade is a corresponding change in the geometry of D1 structural fabrics. We propose that theNorthern Range exposes two distinct tectonic packages that reflect vertical strain partitioning ofthe lithosphere in this Neogene transpressional orogen. Rocks in the eastern (upper) package have relatively low deformation temperatures (200-330°C), upright and slaty D1 fabrics, and NE-SW strikes (Weber et al. in review). Rocks withrelatively low deformation temperatures and identical D1 fabrics also occur in down-droppedfault blocks along the southern foot of the Northern Range (Weber et al. in review) andprobably belong to the same upper tectonic package. Structurally underlying this upper crustalpackage are rocks that are pervasively transposed, highly strained, have been ductileydeformed at higher temperatures (up to 400°C), and contain recumbent D1 structures (e.g.,subhorizontal S1 foliation) and strong E-W stretching lineations. We interpret the eastern package as an upper crustal, non-eroded patch of the fold-and-thrustbelt, which together with widely spaced strike-slip faults dominates the surface structure incentral and southern Trinidad. We suggest that the highly and ductiley strained lower packagerepresents an exhumed, mid-crustal, subhorizontal, detachment zone that separated continuousflow in the mantle (e.g., Russo et al. 1997) from highly partitioned, block-style, deformation(i.e., folds and thrusts with strike-slip faults) above. This upright low-grade over flat high-grade structure, with lineations parallel to the belt, isidentical to the suprastructure-infrastructure architecture in the Pyrenees, which was firstdescribed by De Sitter and Zwart (1960). Drawing on the Pyrenees analogue, we speculatethat either a flat detachment or a complex transitional zone may separate the upper package fromthe lower package in the Northern Range. The origin of this architecture, which may becommon in other transpressional belts, probably reflects vertical strain partitioning in thelithosphere. Flat detachments form in the weak, jelly-like, mid-crust to separate continuousflow in the mantle lithosphere from highly partitioned, block-style, deformation in the uppercrust.


118

CARIBBEAN PROBABILISTIC SEISMIC HAZARD MODELING IN THECOMPLEX ZONES OF TRANSITION FROM SUBDUCTION TO STRIKE-SLIP: HISPANIOLA CASE STUDY

Chesley R Williams1, Steven Masters1, Guy Morrow1 and Robert Muir-Wood 2

1Risk Management Solutions, Inc, Menlo Park, CA 940252RMS Ltd, Peninsular House, 30-36 Monument Street, London EC3R 8LJ UK

The islands of the Caribbean experience a remarkably high concentration of natural hazards, includinghurricanes and earthquakes. To aid insurers and reinsurers in managing their exposure to earthquake- relatedcatastrophes in this region, a probabilistic seismic hazard model for the Caribbean has been developed. Themodel coverage extends from the Cayman Islands through the Northern Antilles to the west, through the LesserAntilles and south to Trinidad and Tobago. The model is constructed from a series of seismic sources for whichevent probabilities and magnitudes have been defined. These seismic sources were delineated through theincorporation of island-specific models and paleoseismic studies, as well as detailed re-examination of historicobservations and seismic catalogs.

According to this new model, the highest ground motions are observed in northern Hispaniola, where the475-year-return-period MM intensity (10% exceedance in 50 years) is slightly greater than IX. The next highesthazard is observed in the Lesser Antilles on Antigua, Barbuda and Guadeloupe, where the 475 intensities are justslightly below IX. The lowest hazard (low VII) at 475 years is in the southern Lesser Antilles (St. Lucia southto Grenada) and along the southern coast of Puerto Rico.

Hispaniola represented one of the most complicated regions to model, since it involves a combination ofthrusting under the northern coast (driven by convergence across the plate boundary) and numerous shallowcrustal fault zones. To cover all of the possible seismic sources in Hispaniola, the final source model utilizesnine shallow surface faults and eight regional sources (Figure 1). Maximum magnitudes and return periods forthe maximum events are shown in the table on the following page.

HS Hatillo FZ

HS Bonao FZ

HS SouthernSamana Bay FZ

HS Hispaniola FZ

HS Septentrional Fault WestHS Camu Fault

HS Septentrional Fault East

HS Enriquillo-Plantation Garden FaultHS Beata FZ

HS Southwest Peninsula

HS SouthCentral Hispaniola

HS Southeast Coast

HS Punta Cana

HS CordilleraCentral

HS Northeast CoastHS North Central CoastCU Southeast Coast

Figure 1. Seismic sources used in the Hispaniola component of the seismic hazard model.


119

Seismic Source Name Max Magnitude Max Return Period Source TypeHS Beata FZ 7.40 3,000 faultHS Bonao FZ 7.25 3,600 faultHS Cordillera Central FZ 7.30 2,500 faultHS Enriquillo-Plantation Garden FZ 7.80 800 faultHS Hatillo FZ 7.10 3,600 faultHS Hispaniola FZ 7.30 3,600 faultHS Septentrional FZ East 8.00 800 faultHS Septentrional FZ West 8.00 800 faultHS South Samana Bay FZ 7.50 3,600 faultHS Camu Fault 7.50 2,400 regionHS North Central Coast 8.30 1,300 regionHS Northeast Coast 8.00 1,000 regionHS Punta Cana Deep 7.80 233 regionHS South Central Hispaniola 7.60 333 regionHS Southeast Coast 7.80 675 regionHS Southwest Peninsula 7.60 333 regionCU Southeast Coast 7.50 150 region

Seismic source locations were delineated from tectonic regions, geologic maps, historical event rupture zonesand microseismicity. During the development of the seismic source recurrence parameters, two critical issuesarose:

1) How should the seismicity between the northern and southern coast sources be distributed?2) How should the seismic moment release between the deeper plate boundary sources and the shallowcrustal faults be partitioned? For this model, roughly two thirds of the seismic moment was assumed to be released along the northern

coast of the island. Of that, about 50% was attributed to the shallow crustal faults and the other 50% wasassociated with the deeper plate boundary sources dipping under the coast. Along the southern boundary of theisland, the Enriquillo-Plantation Garden fault zone was assumed to play a very important role and was given aslip rate just slightly less than that of the Septentrional fault zone. The faults that step over from thesoutheastern coast to the plate boundary in the northwest have been given very low rates. The CordilleraCentral source models the diffuse moderate magnitude events observed in the mountain region. Finally, thePunta Cana source accounts for the large intermediate depth events that occur under eastern Hispaniola.

The answers to the questions above have a critical effect on the resulting probabilistic seismic hazard model.More data to answer these questions are starting to become available through geodetic and paleoseismic studies,and we plan to update our model accordingly.

Important Hispaniola Model References:Dolan, J. F., and Wald, D. J., 1998, The 1943-1953 north-central Caribbean earthquakes: Active tectonic

setting, seismic hazards, and implications for Caribbean-North America plate motions, in Dolan, J. F., andMann, P., eds., Active Tectonics of the Northern Caribbean Plate Boundary Zone, GSA Special Paper 326,p. 143-169.

Farrat, M. R., 1995, Estímaciones Probabilísticas de la Peligrosidad Sismica en Cuba, CENAIS-MAPFRE,BECA 1994-1995, 72 pp.

Mann, P., Draper, G., and Lewis, J. F., 1991, An overview of the geologic and tectonic development ofHispaniola, in Mann, P., Draper, G., and Lewis, J. F., eds., Geologic and Tectonic Development of theNorth America-Caribbean plate boundary in Hispaniola, GSA Special Paper 262, p. 1-28.

McCann, W. R., and Associates, Inc., 1994, Seismic Hazard Map for Puerto Rico 1994, Final DocumentPrepared for the Seismic Safety Commission of Puerto Rico, 60 pp.

Prentice, C. S., Mann, P., Taylor, F. W., Burr, G., and Valastro, S., Jr., 1993, Paleoseismicity of the NorthAmerican-Caribbean plate boundary (Sept. fault), Dominican Republic: Geology, v. 21, p. 49-52.


120

INDEX OF AUTHORS

—A—Aalto, 1Acosta, 24, 113Anton, 24Audemard, 35Audru, 3Avé Lallemant, 21

—B—Barnes, 3Beavan, 23Berryman, 6Beutner, 89Bilich, 8, 31Bodin, 89Bradley, 48Burr, 10, 87Bursik, 46

—C—Calais, 12, 56Chemenda, 50Chiesa, 16, 17Civelli, 16Coleman, 19Collot, 3Comstock, 21

—D—Darby, 23Davy, 3De Toni, 16Delteil, 3Demets, 56Dillon, 24Dixon, 56Dolan, 36, 76, 109Dominguez, 52Doser, 26Draper, 27Dyer-Williams, 107

—E—Erikson, 28

—F—Font, 52Frohlich, 8, 31, 60Fuchs, 42

—G—Gibbons, 33, 79Giraldo, 35Girbacea, 102, 104Grindlay, 36, 76, 109Gulick, 74

—H—Herridge, 38Herzer, 3Hettel, 42, 102

—J—Jansma, 56

—K—Kurbatov, 46Kusky, 48

—L—Lallemand, 50, 52Lamarche, 3Langridge, 54Lebrun, 3Lee, 24Lewis, 55Liu, 52Lomas, 26López, 56Lorenz, 57, 102

—M—Malavieille, 52Mann, 8, 10, 31, 36, 56, 60, 76, 87, 109Martin, 57Masters, 118Mattietti-Kysar, 63Mattioli, 56Mazzoleni, 17McCaffrey, 65McCalpin, 66, 68McCann, 70McCrory, 72McGeehin, 10McGinty, 93Melekestsev, 46Meltzer, 74Mendoza, 115Mercier de Lepinay, 3Merritts, 89Miller, 76


121

Mocanu, 57Molzer, 106Montes, 78Moreno, 79Morrow, 118Moser, 102, 104Moya, 81Muir-Wood, 118Muller, 42, 89Murray, 72

—O—Odonel Gomez, 115

—P—Pena, 107Peña, 87Perez De Armas, 83Pindell, 85Prentice, 10, 87, 89, 107

—R—Ratschbacher, 91, 102, 104Reyners, 93, 95Robinson, 93

—S—Schill, 89Schnürle, 52Scholl, 97Shimizu, 106Silver, 99Sperner, 42, 57, 91, 102, 104Stevens, 99

Sulerzhitsky, 46Sutherland, 3

—T—ten Brink, 24, 106Tuttle, 10, 107

—U—Uchupi, 24

—V—vanGestel, 109Vázquez, 111, 113Vegas, 111, 113Velasquez, 26Velleiux, 26von Hillebrandt-Andrade, 115

—W—Wallace, 99Weber, 117WELDON, 54Wenzel, 42, 57, 102Williams, 118Wilson, 72Winslow, 19Wood, 3

—Z—Zirbes, 115Zweigel, 102, 104

Documents

SUBDUCTION TO STRIKE-SLIP TRANSITIONS ON PLATE BOUNDARIESpeople.uncw.edu/grindlayn/revabstr_vol.pdf · Penrose Conference: Subduction to Strike-Slip Transitions on Plate Boundaries,